Population statistics at regional level

Data extracted in March 2025.

Planned article update: September 2026.

Highlights

On 1 January 2024, Mayotte – 1 of the French outermost regions – had the highest young-age dependency ratio in the EU (123.4%); the Belgian coastal region of Arr. Veurne had the highest old-age dependency ratio (72.8%).

Across the EU, the Spanish capital region of Comunidad de Madrid had the highest life expectancy at birth, 86.1 years in 2023; the north-western Bulgarian region of Severozapaden had the lowest life expectancy at birth, at 73.9 years.

Source: Eurostat (demo_r_pjangrp3)

On 1 January 2024, there were 449.3 million people living in the EU. During the course of 2023, the population of the EU increased by 1.6 million. The rising number of inhabitants resulted exclusively from migratory flows, as natural population change was negative, with 1.2 million more deaths than births. The EU’s net migration – the difference between the number of immigrants and the number of emigrants, with statistical adjustments – was 2.8 million people in 2023. This figure reflects, among other factors, an influx of displaced people linked to Russia’s war of aggression against Ukraine, as well as migrant arrivals for employment, international protection, or family reasons.

Population events such as births, deaths and migratory flows shape demographic changes over time to impact the structure of the EU’s population. Demographic developments are also impacted by irregular shocks, such as the COVID-19 crisis or Russia’s war of aggression against Ukraine. The population pyramids shown in the infographic above highlight the considerable difference in age structures across NUTS level 3 regions. On 1 January 2024, the French outermost region of Mayotte had the highest young-age dependency ratio in the EU, while the Belgian coastal region of Arr. Veurne had the highest old-age dependency ratio.

Statistics based on the population grid

More about the data: statistics on the population grid

Unlike the majority of data presented in the Eurostat regional yearbook, the information presented in the 1st section of this chapter is based on a fixed set of uniform, grid cells that measure 1 km². In the remainder of this publication, data are presented using the more aggregated NUTS 2024 classification. Some of the differences observed when viewing regional data presented by NUTS may reflect the underlying criteria used to determine the administrative boundaries that delineate each region, thereby potentially overlooking important intra-regional disparities and limiting analyses.

The gridded data presented in this section cover the whole of the EU’s territory at a resolution of 1 km². Gridded statistics offer greater spatial precision than regional data, enabling more detailed, localised analyses, while capturing variations within regions that might otherwise be averaged out over larger administrative units. This is particularly valuable when examining high-resolution phenomena, such as population density, land use or environmental factors.

The analyses that follow present gridded statistics based on a dataset from the 2021 population and housing census. The data references a system of uniform, equal-sized grid cells, offering several advantages:

analysts can aggregate the data for different areas that cross regional and national boundaries
the data can be assembled to form areas for specific purposes or studies, such as coastal regions, mountain regions or water catchment areas
the grid system integrates easily with other datasets
unlike administrative classifications, grids remain stable over time.

There were 4.45 million grid cells of 1 km² used for the 2021 population and housing census in the EU, with the EU’s population during this census year equal to 443.2 million inhabitants/residents. Across the EU, there were 1.82 million populated grid cells in the reference grid, with an additional 75 673 populated cells in 3 EFTA countries (Liechtenstein, Norway and Switzerland).

Europeans tend to live in quite densely populated cities or in towns and suburbs, while the vast majority of the territory is more sparsely populated (see Map 1). Almost 60% of the 1 km² grid cells in the EU and EFTA countries – 2.69 million cells – had nobody living in them; among EU countries, Spain and Sweden had the largest areas of uninhabited land (374 735 and 346 556 cells, respectively).

Data from the 2021 housing and population census indicate that 48.9 million people lived in the EU’s capital cities; this represented close to 10% of the overall population. More than 1 in 5 of these capital city residents lived in Paris (France) – the EU’s largest city – where the population was roughly twice the size of the next largest capital, Madrid (Spain).

Looking in detail at individual 1 km² grid cells, some of the most densely populated areas of the EU were located in and around Barcelona (Spain), with a peak to the south-west of the city centre in L'Hospitalet de Llobregat, where 1 grid cell had a population of 56 158 inhabitants. There were also very high population density figures – upwards of 40 000 inhabitants per km² – in some grid cells within Paris and in the Spanish cities of Madrid, Valencia/València and Málaga.

Map 1: Population density
(number of inhabitants, 2021)
Source: Eurostat (GISCO) census population grid 2021

Demographic developments across the EU are far from uniform, with considerable variations both between and within individual countries. The EU’s population structure is evolving, with a key factor being the increased mobility of young people. The EU’s population distribution spans various settlement types, each characterised by distinct demographic and economic features, including:

major urban areas that tend to attract younger populations, with a high share of people living alone, relatively high living costs, diverse educational opportunities and dynamic labour markets
commuter belts/suburban areas, which are often populated by families
former industrial heartlands, characterised by economic decline, lower living costs, relatively high levels of unemployment, poverty and social exclusion
coastal and countryside locations, some of which are viewed as retirement locations for relatively affluent pensioners
other rural and remote regions which frequently exhibit declining population numbers, ageing populations, limited job opportunities and lower access to essential services.

Population developments can be analysed using data from the 2011 and 2021 population and housing censuses. According to this source, Germany had the largest absolute increase in population, with more than 1.7 million additional inhabitants. In relative terms, Malta experienced the fastest population growth, up 24.5%. By contrast, Romania recorded the largest population decline, losing more than a million inhabitants during the period under consideration, while Bulgaria recorded the steepest fall in relative terms, down 11.5%.

Between 2011 and 2021, several grid cells saw a significant increase in their population counts, with more than 10 000 additional inhabitants per 1 km². They were located in:

several different urban areas within Byen København (the Danish capital)
Malmö, in the southern Swedish region of Skåne län (which is at the opposite end of the Öresund Bridge that links it to København)
Aalborg, in the northern Danish region of Nordjylland
3 other capital cities: Miasto Warszawa (Poland), Bucureşti/Ilfov (Romania) and Stockholms län (Sweden)
3 French cities: Boulogne-sur-Mer (Pas-de-Calais), Marseille (Bouches-du-Rhône) and Toulon (Var).

Map 2: Population change
(overall change in number of inhabitants, 2011–21)
Source: Eurostat (GISCO) census population grid 2021

Population structure

Historically, life expectancy at birth in the EU has increased at a relatively consistent pace. In 2019, prior to the COVID-19 pandemic, life expectancy at birth had reached 81.3 years. However, it fell by 0.9 years in 2020 and by an additional 0.3 years in 2021. In 2022, life expectancy at birth began to rise again, and this upward pattern continued in 2023, reaching 81.4 years, which was 0.1 years higher than its pre-pandemic level.

There are a number of drivers that impact inter-regional differences in life expectancy, including:

proximity to healthcare services – for example, capital regions tend to have a greater number and variety of healthcare facilities compared with rural regions
the prosperity of a region – for example, life expectancy is generally above average in regions characterised by a higher standard of living and below average in regions characterised by poverty and social deprivation
lifestyle and cultural differences – for example, the type of work that predominates in a region, the typical diet of a region, or the incidence of smoking and alcohol consumption
climatic conditions – for example, people living in warm and relatively dry climates tend to live longer lives than those living in regions that experience more extreme weather conditions.

Map 3 is composed of 2 parts: it presents life expectancy at birth for females and for males. Both maps use the same class boundaries in their legends to facilitate comparison. In 2023, female life expectancy at birth in the EU was 84.0 years. This equated to a gap of 5.3 years between the sexes, as male life expectancy at birth was 78.7 years.

Female life expectancy was higher than male life expectancy across every region of the EU

In 2023, all 242 NUTS level 2 regions for which data are available had higher female (than male) life expectancy at birth. The Baltic countries, along with several regions in Poland and Romania, reported some of the largest gender gaps. In Latvia, female life expectancy at birth (80.6 years) exceeded that for males (70.5 years) by 10.1 years.

At the other end of the range, all of the regions in the Netherlands (subject to data availability) and all but 1 of the regions in Sweden – the northernmost region of Övre Norrland being the only exception – reported much smaller differences between the sexes. Flevoland in the Netherlands had the narrowest regional gender gap in 2023, with female life expectancy at birth (83.3 years) exceeding that for males (80.8 years) by 2.5 years. Several regions outside of the Netherlands and Sweden also reported gender gaps that were no more than 3.5 years:

Prov. Antwerpen in Belgium
Nordjylland in Denmark
Eastern and Midland, and Southern in Ireland
Ciudad de Melilla in Spain
Mayotte in France
Luxembourg.

In 2023, the highest levels of female life expectancy at birth (as shown by the darkest shade of blue in Map 3) were located across much of Spain and Italy. There were also relatively high values – at least 85.5 years – recorded in several regions from Belgium, Greece, France, Austria and Portugal, as well as single regions from Slovenia, Finland and Sweden. The Finnish archipelago of Åland had the highest female life expectancy at birth, peaking at 88.5 years. The next highest regional values were concentrated in central and northern Spain: Comunidad de Madrid (88.3 years), Castilla y León (87.7 years), Comunidad Foral de Navarra (87.6 years) and País Vasco (87.4 years).

Several of the highest levels of male life expectancy at birth were concentrated in northern and central regions of Italy. Several regions in Belgium, Spain, France and Sweden, as well as Åland (Finland), Eastern and Midland (the Irish capital region), Luxembourg and Malta also recorded relatively high values – at least 81.5 years. Comunidad de Madrid had the highest male life expectancy at birth (83.4 years in 2023). After the Spanish capital region, the next highest values were in Provincia Autonoma di Bolzano/Bozen, Provincia Autonoma di Trento (both northern Italy) and Stockholm (the Swedish capital region), each with male life expectancy at birth of 82.8 years.

Across the 242 EU regions for which data are available, the French outermost region of Mayotte recorded the lowest female life expectancy at birth (76.2 years in 2023). There were 2 other regions – Severozapaden in north-western Bulgaria and Észak-Magyarország in northern Hungary – which also recorded female life expectancy at birth below 79.0 years, as shown by the lightest 2 shades in Map 3.

There were 49 NUTS level 2 regions across the EU where male life expectancy at birth was below 76.0 years in 2023 (as indicated by the lightest shade in Map 3). These regions were concentrated in eastern EU countries – Bulgaria, Czechia, Croatia, Hungary, Poland, Romania and Slovakia – but also included all of the regions in the Baltic countries, as well as Mayotte. Severozapaden (70.0 years) and Latvia (70.5 years) recorded the lowest regional values for male life expectancy at birth.

Map 3: Life expectancy at birth
(years, by NUTS 2 regions, 2023)
Source: Eurostat (demo_r_mlifexp) and (demo_mlexpec)

The median age is an indicator that helps summarise the pace at which population structures have changed. During the last 2 decades, the median age of the EU population increased 5.4 years, up from 39.3 years on 1 January 2004 to 44.7 years by 1 January 2024.

The regional distribution of median ages exhibits some skewness. On 1 January 2024, 744 NUTS level 3 regions recorded a median age that was above the level for the EU, 13 had a median age that was the same as the EU (44.7 years), while 408 regions had median ages below the level for the EU.

At the upper end of the distribution, 178 regions in the EU had a median age of at least 50.0 years. Germany accounted for 79 regions within this group (nearly half of the 178 regions), while there were 40 regions located in Italy (almost 1 in 4 of the 178 regions). Most of the remaining regions with relatively high median ages were located in Bulgaria, Greece, France and Portugal (each with 10 regions).
At the lower end of the distribution, 50 regions across the EU had a median age below 40.0 years. France (14 regions) and Germany (11 regions) together accounted for half of this group. Many of the regions with low median ages were predominantly urban regions (including capital regions and regions that border/surround their capital).

At the start of 2024, rural regions recorded some of the highest median ages …

Figure 1 highlights those regions with the highest and lowest median ages, as of 1 January 2024. There were 3 regions in the EU that reported a median age above 55.0 years:

Evrytania (57.0 years), a mountainous region in central Greece
Arr. Veurne (56.6 years), a rural/coastal region in western Flanders (Belgium)
Alto Tâmega e Barroso (56.5 years), a mountainous region in northern Portugal.

… while some of the lowest median ages were in and around capital cities

Capital regions often exert a considerable pull on inter-regional and international migrants, as they tend to provide a diverse range of educational and employment opportunities. This process can lead to a shift in population structures, with younger people accounting for a growing share of the population in predominantly urban regions; over time, this pattern may self-reinforce, insofar as populations with younger age structures are more likely to have relatively high birth rates.

As of 1 January 2024, the outermost regions of Mayotte and Guyane (both France) had the lowest median ages in the EU (18.1 years and 27.3 years, respectively), reflecting high fertility rates and lower life expectancy; this was also the case in the autonomous region of Melilla (Spain). Beyond these 3 atypical regions, the lowest median ages were recorded in:

capital regions, including the Danish capital of Byen København (34.1 years), the Belgian capital of Arr. de Bruxelles-Capitale/Arr. Brussel-Hoofdstad (36.0 years) and the Irish capital of Dublin (37.8 years)
regions bordering or surrounding capitals, such as Seine-Saint-Denis and Val-d’Oise in France
other predominantly urban regions that are home to renowned academic institutions, diverse populations and dynamic labour markets, like Heidelburg and Freiburg im Breisgau (both Germany).

Figure 1: Median age
Source: Eurostat (demo_r_pjanind3) and (demo_r_pjanind)

As of 1 January 2024, the Belgian coastal region of Arr. Veurne had the highest old-age dependency ratio in the EU

Europe has an ageing population: dependency ratios serve as a key measure for analysing the economic and social pressures associated with low birth rates, a shrinking workforce and a growing number of retirees, driven in part by increased longevity. For the purpose of this publication, the old-age dependency ratio is defined as the proportion of elderly individuals (aged 65 years or over) relative to the working-age population (people aged 20 to 64 years); it is expressed in percentage terms.

With a growing share of elderly people in the population, the EU faces a number of challenges in relation to pension systems, healthcare and labour markets, emphasising the need for sustainable policies that support the ageing population while at the same time ensuring economic and social stability.

During the past 2 decades, the EU’s old-age dependency ratio has increased at a rapid pace. On 1 January 2004, it stood at 26.8%, indicating that there were slightly fewer than 4 working-age adults for every elderly person aged 65 years or over. Fast-forward to 1 January 2024 and the ratio had risen to 37.0%, meaning there were fewer than 3 working-age adults per elderly person.

As of 1 January 2024, there were 139 NUTS level 3 regions where the old-age dependency ratio was at least 50.0%, indicating that there were no more than 2 working-age adults for every elderly person. This group of 139 regions, shown by the darkest shade of blue in Map 4, were mainly concentrated in (eastern) Germany and France. In addition, Italy, Finland, Portugal, Bulgaria, Greece and Spain also had several regions where the old-age dependency ratio was at least 50.0%. Most of these regions with high old-age dependency ratios were characterised as predominantly rural, mountainous and/or relatively remote regions. In other words, regions where it is likely that younger people have left the region to continue their studies or look for jobs.

Arr. Veurne, a coastal region in the west of Belgium had the highest old-age dependency ratio in the EU, at 72.8% on 1 January 2024.
Alto Tâmega e Barroso in northern Portugal and Evrytania in central Greece were the only other regions in the EU where this ratio exceeded 70.0%.

Predominantly urban regions generally provide more opportunities for higher education and employment across a diverse range of occupations. As a result, they often attract young people, leading to lower old-age dependency ratios. Equally, some elderly people may choose to leave these urban regions once they reach retirement, to avoid some of their perceived disadvantages, such as congestion, noise, crime and/or higher living costs.

Leaving aside the atypical outermost regions of Mayotte (6.1%) and Guyane (13.8%), on 1 January 2024, the lowest old-age dependency ratios were generally recorded in some of the EU’s most dynamic and diverse regions, including:

capital regions – Byen København in Denmark (17.8%), Arr. de Bruxelles-Capitale / Arr. Brussel-Hoofdstad in Belgium (20.5%), Dublin in Ireland (22.0%), Luxembourg (23.5%) and Groot-Amsterdam in the Netherlands (24.4%)
regions bordering or surrounding capital regions – Ilfov in Romania (21.0%), Seine-Saint-Denis in France (22.6%), Mid-East in Ireland (23.8%)
other predominantly urban regions – the German regions of Frankfurt am Main (24.3%), its neighbouring region of Offenbach am Main (25.0%) and Heidelberg (25.0%), as well as Gdański in Poland (24.7%).

Map 4: Old-age dependency ratio
Source: Eurostat (demo_r_pjanind3) and (demo_pjanind)

Figure 2 presents a detailed analysis of population structures across NUTS level 3 regions. It highlights those regions with the highest and lowest shares of their populations across 3 different age groups, as of 1 January 2024.

Young people (less than 20 years) made up 20.0% of the EU’s total population.

Notably, there were 2 regions that had shares of young people that were more than twice as high as across the whole of the EU: the French outermost regions of Mayotte (53.8%) and Guyane (40.2%).
Several predominantly rural regions in southern EU countries recorded particularly low shares of young people: the lowest shares were in the north-western Spanish region of Zamora (12.7%), the northern Portuguese region of Alto Tâmega e Barroso (12.8%) and the central Greek region of Evrytania (12.9%).

Working-age people made up 58.4% of the EU’s total population.

The highest regional share was recorded in the Spanish island region of Eivissa y Formentera, where 68.9% of the total population was of working age. Other regions with relatively high shares included:
- tourism-driven economies – Fuerteventura, Lanzarote and Gran Canaria (all in Spain), and Malta
- urban/metropolitan hubs – Byen København (Denmark) and Frankfurt am Main (Germany)
- university/research centres – Heidelberg and Regensburg (both in Germany).
There were 4 regions in the EU where less than half of the population was of working age:
- Mayotte (France) – where a relatively high share of the population is composed of young people and some individuals seek to migrate for education and/or employment opportunities
- Etelä-Savo (Finland), Arr. Veurne (Belgium) and Lot (France) – all of which attract retirees seeking a calmer lifestyle centred on their natural surroundings.

Older people accounted for 21.6% of the EU’s total population.

There were 7 regions in the EU where at least 1 in 3 of the total population was composed of older people, several of them facing issues linked to depopulation. The highest shares were in:
- the mountainous, inland Portuguese regions of Alto Tâmega e Barroso (which recorded the highest share, at 36.3%), Terras de Trás-os-Montes, Beiras e Serra da Estrela and Beira Baixa
- the mountainous Greek region of Evrytania
- the coastal Belgian region of Arr. Veurne
- the lakeland Finnish region of Etelä-Savo.
By contrast, older people accounted for less than a tenth of the population in the French outermost regions of Mayotte (2.7%) and Guyane (7.3%). There were also relatively low shares in:
- Melilla and Ceuta (both in Spain)
- urban/metropolitan hubs – Byen København (Denmark), Arr. de Bruxelles-Capitale/Arr. Brussel-Hoofdstad (Belgium), Ilfov (Romania) and Seine-Saint-Denis (France)
- tourism-driven economies – Fuerteventura and Eivissa y Formentera (both in Spain).

Figure 2: Younger, working-age and older people
Source: Eurostat (demo_r_pjanind3) and (demo_r_pjangroup)

Fertility and infant mortality

More about the data: measuring fertility

The total fertility rate is the average (mean) number of children who would be born to a woman during her lifetime, if she were to spend her childbearing years conforming to the age-specific fertility rates of a given year.

The natural replacement rate – the average number of live births per woman required to keep the population size constant in the absence of migration – is estimated to be around 2.10 children per woman for developed countries.

Since 2016, the number of live births across the EU declined slowly from 4.4 million in 2016, falling over 4 consecutive years. After a modest rebound in 2021, the pace of decline accelerated in 2022 and 2023, with the number of births falling 5.1% and 5.4%, respectively. By 2023, the number of live births stood at 3.7 million; this was equivalent to an overall decrease of 16.3% compared with 2016.

In 2023, the EU’s total fertility rate was 1.38 live births per woman; this was the lowest rate recorded since the start of the time series in 2001. During the last 2 decades, the EU’s fertility rate has remained consistently below the natural replacement rate, within the range of 1.38 to 1.57 live births per woman.

French outermost regions have some of the highest fertility rates in the EU

Map 5 shows the distribution of total fertility rates across NUTS level 3 regions. In 2023, the regional distribution was somewhat skewed insofar as there were 423 regions (38.7% of all regions) where the total fertility rate was below the EU level, while there were 671 regions (61.3%) where the rate was equal to or above the EU level.

A more detailed analysis of the latest regional data for 2023 reveals that:

some of the highest total fertility rates were recorded across Bulgaria and France
there were 6 regions across the EU where the total fertility rate was higher than the natural replacement rate of 2.10 live births per woman (as shown by the darkest shade of blue in Map 5)
- the French outermost regions of Mayotte (4.17 live births per woman), Guyane (3.55) and La Réunion (2.31)
- the Bulgarian regions of Sliven (2.58), Yambol (2.40) and Pazardzhik (2.20)
some of the lowest total fertility rates, below 1.00 live birth per woman, were concentrated in the southern EU countries of Greece, Spain and Italy (as shown by the lightest shade of yellow in Map 5)
- the bottom end of the distribution had 5 island regions from Canarias (Spain) – El Hierro recorded the lowest rate (at 0.75 live births per woman), while rates were only marginally higher in Tenerife, La Gomera, Gran Canaria and La Palma
- leaving aside these Spanish island regions, the next lowest rates occurred on the Italian island region of Sardegna – in Cagliari and Sud Sardegna – and in the central Greek region of Fokida
it was relatively common for the lowest total fertility rate within a country to be recorded in the capital region; this was the case for Hlavní město Praha (Czechia), Byen København (Denmark), Põhja-Eesti (Estonia), Dublin (Ireland), Budapest (Hungary), Malta (Malta) and Wien (Austria).

Map 5: Total fertility rate
Source: Eurostat (demo_r_find3) and (demo_find)

Women in the EU are giving birth later in life

A growing proportion of women in the EU give birth later in life, explaining, at least in part, the relatively low levels of fertility. This may be linked, among other factors, to:

higher female participation rates in further education and/or more women choosing to establish a career before starting a family
lower levels of job security (for example, in precarious employment)
the increasing cost of raising children and of housing.

More about the data: the mean and the median age of women at childbirth

The mean age of women at childbirth is the average age of mothers when giving birth; it is calculated by summing up all of the mothers’ ages at childbirth and dividing by the total number of births. The mean value can be skewed by extreme values, such as when a few women give birth much earlier or much later in life.

The median age at childbirth is the middle value when all the mothers’ ages at childbirth are arranged in order. It usually provides a more reliable measure of the ‘typical age’ at which mothers give birth, insofar as it is less affected by outliers.

Both measures provide insights into fertility patterns across the EU. If the mean age is lower than the median age, the distribution is negatively skewed, with some younger mothers – teenagers and women in their early 20s – pulling the mean down. If the mean age is higher than the median age, the distribution is positively skewed, with most births concentrated at younger or middle reproductive ages, but some older mothers raising the mean.

Across the EU, the median age of women at childbirth rose from 30.8 to 31.8 years between 2013 and 2023. During this decade, the rise in the median age had a fairly regular pattern of development, with modest annual increases interspersed with occasional years of stability.

In 2023, the predominantly urban region of Voreios Tomeas Athinon, located to the north of the Greek capital, recorded the highest median age for mothers at childbirth …

Figure 3 (right-hand side) shows that in 2023, among NUTS level 3 regions:

some of the highest median ages of mothers at childbirth occurred in and around the Greek capital: Voreios Tomeas Athinon (to the north; 35.6 years, the highest value across the EU), Notios Tomeas Athinon (to the south; 34.7 years) and Kentrikos Tomeas Athinon (the capital region; 34.3 years)
several regions in the north and north-west of Spain had high median ages, including: A Coruña (34.8 years), Pontevedra (34.5 years), Bizkaia (34.4 years), Lugo, Ourense, Asturias, Araba/Álava and Valladolid (all 34.3 years)
3 regions on the Italian island of Sardegna had relatively high median ages: Oristano (34.7 years), Cagliari (34.4 years) and Sud Sardegna (34.3 years)
2 predominantly rural regions in Ireland, West and South-West, had relatively high median ages (34.9 years and 34.6 years, respectively), as did the capital region of Dublin (34.3 years).

At the bottom end of the distribution, the 10 EU regions with the lowest median ages of mothers at childbirth in 2023 were equally split between Bulgaria and Romania (5 regions each). The lowest values occurred in:

the south-eastern Bulgarian regions of Sliven (23.5 years) and Yambol (25.1 years), both of which had total fertility rates higher than the natural replacement rate
the southern Romanian regions of Călăraşi (26.0 years) and Ialomiţa (26.2 years).

Between 2001 and 2023, the mean age of women at childbirth in the EU steadily increased from 29.0 to 31.2 years. Over the last 2 decades, this indicator followed a regular upward pattern of development, with modest annual increments or occasional years of stability. By 2023, the mean age of mothers at childbirth stood at 31.2 years.

Capital regions and predominantly urban regions accounted for some of the highest mean ages of mothers at childbirth. This may reflect a variety of cultural, socioeconomic and personal factors, including:

better access to healthcare in these regions, increasing the likelihood of a successful pregnancy for older women
more progressive social norms, giving women more freedom to decide when they have children.

… while Voreios Tomeas Athinon also had the highest mean age for mothers at childbirth

Figure 3 (left-hand side) highlights the NUTS level 3 regions with the highest and lowest mean ages of mothers at childbirth. In 2023, the Greek region of Voreios Tomeas Athinon recorded the highest mean age (it also had the highest median age). Several other predominantly urban regions in and around the Greek capital appeared near the top of the ranking, along with Paris in France (34.3 years) and Byen København in Denmark (33.8 years).

At the opposite end of the distribution, there were 7 EU regions where the mean age of women at childbirth was below 27.0 years; they were all located in Bulgaria (5 regions) or Romania (2 regions). The lowest mean ages were recorded in 3 Bulgarian regions – Sliven (25.0 years), Yambol (25.8 years) and Pazardzhik (26.4 years) – all of which had total fertility rates higher than the natural replacement rate.

Figure 3: Mean and median age of mothers at childbirth, 2023
Source: Eurostat (demo_r_find3) and (demo_find)

Within the EU, a marked reduction in infant mortality rates has been a key driver of increased life expectancy. The EU has a relatively low infant mortality rate by international standards, reflecting well-established healthcare systems, access to quality prenatal and neonatal care, and comprehensive social support.

In 1961, the EU recorded an infant mortality rate of 38.2 deaths per 1 000 live births. This rate fell almost uninterrupted and at a relatively fast pace through to 2000, when it stood at 6.0 deaths per 1 000 live births. The downward trend continued, albeit at a slower pace, over the next 2 decades. In 2023, some 12 280 children across the EU died before reaching the age of 1; as such, the infant mortality rate was 3.3 deaths per 1 000 live births.

Capital regions have some of the lowest infant mortality rates

Map 6 shows that the distribution of infant mortality rates across NUTS level 2 regions in 2023 was somewhat skewed. Of the 244 regions for which data are available, 140 had rates below that of the EU and 96 had rates above; the remaining 8 regions had rates identical to that of the EU.

In 2023, 19 regions across the EU had infant mortality rates that were below 2.0 deaths per 1 000 live births (as shown by the lightest shade of yellow in Map 6). These relatively low rates were mainly concentrated in Italy (4 regions), Finland, Sweden (both 3 regions) and Austria (2 regions). A closer look reveals that several of the regions with low infant mortality rates were capital regions. Budapest (Hungary), Helsinki-Uusimaa (Finland), Stockholm (Sweden), Zahodna Slovenija (Slovenia), Praha (Czechia) and Lazio (Italy) all recorded infant mortality rates that were i) below their respective national averages, and ii) below 2.0 deaths per 1 000 live births. These low rates may reflect, among other factors:

higher living standards
better access to healthcare facilities and/or
a concentration of expertise and resources (for example, specialised neonatal units for infants requiring advanced medical interventions).

At the upper end of the distribution, 35 regions had infant mortality rates of at least 4.5 deaths per 1 000 live births in 2023 (as shown by the 3 darkest shades of blue in Map 6). This group included:

all 5 of the French outermost regions – Mayotte (10.5 deaths per 1 000 live births), Guadeloupe (9.7), Guyane (9.6), Martinique (7.5) and La Réunion (5.6) – with Mayotte recording the highest rate in the EU
a large number of predominantly rural regions, such as Východné Slovensko in Slovakia (9.3 deaths per 1 000 live births), Yugoiztochen in Bulgaria (7.8), Nord-Vest in Romania (6.9) and Anatoliki Makedonia, Thraki in Greece (6.6)
some exceptions, as there were a few regions characterised by much higher levels of population density, including Bremen in Germany, Flevoland in the Netherlands, and the capital regions of Wien in Austria and Ile-de-France in France.

Map 6: Infant mortality rate
Source: Eurostat (demo_r_minfind) and (demo_minfind)

Source data for figures and maps

Download Excel file

Data sources

The information presented at the start of this chapter is based on population grids; the data are sourced from the 2021 population and housing census. There were 4 different approaches used to collect these data: traditional, register-based, rolling and combined census. Data for Iceland are not available.

Commission Implementing Regulation (EU) 1799/2018 lays down the criteria for gridded information provided for 1 km² grid cells from the 2021 population and housing census as follows. With a deadline of 31 December 2022:

total population.

With a deadline of 31 March 2024, data by:

sex (males, females)
age (under 15, 15 to 64 years, 65 years or over)
employed people
place of birth (in the reporting country, in another EU country, outside the EU)
usual residence 12 months prior to the census (unchanged, within the reporting country, outside of the reporting country).

The census dataset uses a common reference grid, consistent with the INSPIRE framework. It is defined by Regulation (EU) No 1089/2010, that specifies that the ETRS89-LAEA coordinate reference system should be used. The geo-referenced population dataset from the population and housing census can be accessed via Eurostat’s GISCO website.

Currently, there are 30 countries in the European Statistical System – all 27 EU countries, as well as Liechtenstein, Norway and Switzerland – that have provided geo-referenced, grid-based official data for the 2021 population and housing census. To create their population grids, the EU and EFTA countries – other than Greece and France – applied the aggregation method (or bottom-up approach), using geocoded micro data to assign populations to individual cells of a square kilometre. In Greece and France, the hybrid method was adopted instead – it combines aggregation and disaggregation techniques and represents a compromise between the accuracy and availability of data.

Population census data at a 1 km² grid level have the same scale, the same resolution and the same delineation across countries, in order to allow easy comparisons and combinations/aggregations of information.

The first population grid data available from the 2021 population and housing census was for the total population by place of usual residence. Although this topic is considered non-sensitive, most countries set confidentiality thresholds defining the minimum number of people for each cell that could be published without having to suppress data. The following countries did not set a threshold: Bulgaria, Czechia, Denmark, France, Hungary, Italy and Luxembourg. Finland applied a different method to protect information.

Data for the land surface area for the 1 km² population grid was calculated for 4 countries (Czechia, France, Cyprus and Norway), where the most precise geographical information – for 1 hectare – was not available. Rather, the land surface parameters for these 4 countries are based on the CORINE dataset (minimum mapping unit 25 hectares) and the EuroBoundaryMap (EBM) administrative boundaries.

The dissemination of EU-wide harmonised census topics on a constant area grid, in particular on a 1 km² grid, is a key European statistical output for evidence-based policymaking. The collection of geocoded population data addresses a common need for reliable, accurate and comparable information on population distributions with sufficient spatial resolution.

Regional demographic statistics

Eurostat collects a wide range of regional demographic statistics, including data on population numbers and various demographic events which influence the population’s size, structure and specific characteristics. Regional demographic statistics may be used for a wide range of planning, monitoring and evaluating actions, for example, to:

study population ageing and its effects on sustainability and welfare
evaluate the economic impact of demographic change
calculate ratios relative to the size of the population – for example, regional GDP per inhabitant – which is a key measure when allocating structural funds to economically less advantaged regions.

Demographic statistics include information on:

the usual resident population, which counts the number of people living in a given area on 1 January (or, in some cases, 31 December of the previous year)
the number of live births
the number of deaths.

These statistics are presented for different NUTS levels within the EU, EFTA and candidate countries.

NUTS level 2

Population by sex, age and region of residence
Live births by mother’s age, mother’s year of birth and mother’s region of residence
Deaths by sex, age, year of birth and region of residence
Life table by age and sex
Life expectancy by age and sex
Infant mortality and infant mortality rates

NUTS level 3

Population by sex, 5-year age group and region of residence
Live births by 5-year age group of the mothers and region of residence
Deaths by week, sex, 5-year age group and region of residence
Demographic balance and crude rates (population change, natural change, net migration including statistical adjustment, crude birth and death rates, and crude rates of population change)
Population structure indicators (shares of various population age groups, dependency ratios and median ages)
Fertility indicators (total fertility rate, mean age of woman at childbirth, median age of women at childbirth)
Population density

Regional demography statistics are collected in accordance with Article 3 of Regulation (EU) No 1260/2013 of the European Parliament and of the Council of 20 November 2013 on European demographic statistics and the implementing measures laid out in Commission Implementing Regulation (EU) No 205/2014 of 4 March 2014.

Eurostat has collected regional demographic data according to this legal basis since the reference year 2013. Before 2013, national statistical offices provided regional demographic data on a voluntary basis.

Population

For population statistics, Eurostat recommends the use of the ‘usually resident population’ of a given area. The number of inhabitants is compiled for 1 January of the year in question (or, in some cases, on 31 December of the previous year). Countries can also base their population statistics on the most recent census, adjusted by the components of population change since then, or on population registers.

The average population during a calendar year is calculated as the arithmetic mean of the population on 1 January of the reference year and of the following year. This measure is used, among other purposes, in the calculation of demographic indicators, such as crude rates per 1 000 people.

Population density

Population density is the number of inhabitants per square kilometre (km²). The land area concept (which excludes inland water bodies like lakes or rivers) should be used as the denominator whenever available.

Life expectancy

Life expectancy at birth is the mean number of years a newborn child can expect to live if subjected throughout their life to the current mortality conditions. Life expectancy is normally calculated separately for all age levels, as well as for males, females and the total population.

Median age

The median age is the age that divides a population, ranked by age, into 2 equal-sized groups.

Age dependency ratio

Age dependency ratios typically compare economically dependent people with those who are not. For the purpose of this chapter, the economically dependent population is defined as the sum of people aged less than 20 years and people aged 65 years or over; these 2 groups are generally considered to be economically inactive, either because they are probably still in education or because they are probably retired from the labour force.

By contrast, the economically productive population consists of people of working age, defined here as those aged 20 to 64 years. Age dependency ratios are calculated for each region by dividing the economically dependent population by the number of people of working age; the results are expressed as a percentage. This chapter presents information on the young-age dependency ratio (the number of people aged less than 20 years compared with the number of people aged 20 to 64 years) and the old-age dependency ratio (the number of people aged 65 years or over compared with the number of people aged 20 to 64 years).

Fertility

Fertility is the ability to conceive (become pregnant) and give birth to children. The total fertility rate is defined as the mean number of children who would be born to a woman during her lifetime, if she were to spend her childbearing years conforming to the age-specific fertility rates that have been measured in a given year.

A birth is defined as the start of life when a child emerges from the body of its mother. The total number of births includes both live births and stillbirths (foetal deaths). A live birth is the birth of a child who shows any sign of life. The number of live births refers to the number of births excluding stillbirths.

The median age of women at childbirth is the age that divides the population of mothers at childbirth in 2 numerically equal groups, meaning half of the mothers are younger than the median age and half are older. The mean age of women at childbirth is calculated by weighting the age of mothers at childbirth by the fertility rates for each age group. Both the mean and median age of women at childbirth tend to be relatively high in developed economies.

Infant mortality

A death, according to a United Nations definition, is the ‘permanent disappearance of all vital functions without possibility of resuscitation at any time after a live birth has taken place’; this definition therefore excludes stillbirths. Infant mortality refers to the death of live-born children aged less than 1 year.

The infant mortality rate is defined as the ratio of the number of deaths of children under 1 year of age to the number of live births in the reference year; the value is expressed per 1 000 live births.

Context

A variety of factors, including achievements in medicine, socioeconomic changes and a pattern of increasing urbanisation have shaped demographic developments across the EU. Policymakers can regulate some of the perceived issues linked to demographic transition through actions that prevent, delay, or address demographic imbalances by introducing measures that impact fertility rates, the process of population ageing and/or the flow of migrants (nationally and internationally). These interventions can be direct (for example, vaccination programmes for young children) or indirect (for example, tax breaks or social transfers that provide an incentive for people to have (more) children).

Prolonged life expectancy represents a considerable social achievement. However, when coupled with historically low fertility rates, it has led to a major change in the structure of the EU’s population, with a growing share of elderly people. These developments may pose a range of societal challenges, with a higher proportion of the population considered as unproductive or inactive (those aged 65 years or over), while relatively few younger people enter the labour force. The growing number of very old people in the EU also impacts the sustainability of welfare, health and care systems and may require the development of a broad range of new services to meet the specific demands of an increasingly frail population.

Population ageing is already apparent in several EU countries. It manifests itself through labour market shortages for certain occupations and puts the EU’s long-term economic prosperity and competitiveness at risk. Policymakers can alleviate these challenges to some degree through initiatives that encourage, among other actions, more flexible working opportunities, better provision of childcare, or older people to remain in the labour market for longer.

The European Commission has addressed the on-going demographic transition by adopting:

a Green Paper on Ageing – Fostering solidarity and responsibility between generations (COM(2021) 50 final)
A long-term vision for the EU’s rural areas – Towards stronger, connected, resilient and prosperous rural areas by 2040
a Communication on Harnessing talent in Europe’s regions (COM(2023) 32 final)
a Communication on Demographic change in Europe: a toolbox for action (COM(2023) 577 final).

The EU has been going through a period of rapid demographic and societal change. On 17 January 2023, the European Commission published a Staff Working Document on The impact of demographic change in a changing environment which provided analyses of long-terms demographic trends as well as the specific consequences of one-off events such as Brexit, the COVID-19 pandemic and the Russian war of aggression against Ukraine. The European Commission is seeking to tackle the impact of demographic change through a number of initiatives, including:

a demography toolbox
an atlas on demography
a long-term vision for EU rural areas
a European care strategy
a strategy for the rights of children (including a child guarantee, fighting sexual abuse and better internet for kids)
addressing loneliness in the EU.

This article forms part of Eurostat’s annual flagship publication, the Eurostat regional yearbook.

You can explore the maps interactively using Eurostat’s Statistical Atlas.

Explore further

Database

Regional statistics by NUTS classification – detailed datasets (reg), see

Regional demographic statistics (reg_dem)

Population and area (reg_dempoar)

Fertility (reg_demfer)

Mortality (reg_demmor)

Demography, population stock and balance – detailed datasets (demo), see

Main population indicators (demo_ind)

Population (national level) (demo_pop)

Population (regional level) (demopreg)

Fertility (national level) (demo_fer)

Fertility (regional level) (demofreg)

Mortality (national level) (demo_mor)

Mortality (regional level) (demomreg)

Thematic section

Publications

Paper and PDF publications

Ageing Europe – Looking at the lives of older people in the EU – 2020 edition
Eurostat regional yearbook – 2025 edition
The 2021 population and housing censuses in the EU
Urban Europe – Statistics on cities, towns and suburbs – 2016 edition

Online and interactive publications

Selected datasets

Regional statistics (t_reg), see

Regional demographic statistics (t_reg_dem)

Demography, population stock and balance – selected datasets (t_demo), see

Population (regional level) (t_demopreg)

Population on 1 January by NUTS 2 region (tgs00096)

Population change by NUTS 2 region – Crude rates of total change, natural change and net migration plus adjustment (tgs00099)

Fertility and mortality (regional level) (t_demofmreg)

Total fertility rate by NUTS 2 region (tgs00100)

Life expectancy at birth by sex and NUTS 2 region (tgs00101)

Methodology

Manuals and further methodological information

Metadata

Census 2021 round (ESMS metadata file – cens_21_esms)
Population (ESMS metadata file – demo_pop_esms)
Population change – demographic balance and crude rates at national level (ESMS metadata file – demo_gind_esms)
Population change – demographic balance and crude rates at regional level (NUTS 3) (ESMS metadata file – demo_r_gind3_esms)

External links

Legislation

Regulation (EU) No 1260/2013 of the European Parliament and of the Council of 20 November 2013 on European demographic statistics (Text with EEA relevance)
- Commission Implementing Regulation (EU) No 205/2014 of 4 March 2014 laying down uniformed conditions for the implementation of Regulation (EU) No 1260/2013 of the European Parliament and the Council on European demographic statistics, as regards breakdowns of data, deadlines and data revisions (Text with EEA relevance)
Regulation (EC) No 763/2008 of the European Parliament and of the Council of 9 July 2008 on population and housing censuses (Text with EEA relevance)
- Commission Implementing Regulation (EU) 2018/1799 of 21 November 2018 on the establishment of a temporary direct statistical action for the dissemination of selected topics of the 2021 population and housing census geocoded to a 1 km² grid (Text with EEA relevance)
- Commission Implementing Regulation (EU) 2017/881 of 23 May 2017 implementing Regulation (EC) No 763/2008 of the European Parliament and of the Council on population and housing censuses, as regards the modalities and structure of the quality reports and the technical format for data transmission, and amending Regulation (EU) No 1151/2010 (Text with EEA relevance)
- Commission Regulation (EU) 2017/712 of 20 April 2017 establishing the reference year and the programme of the statistical data and metadata for population and housing censuses provided for by Regulation (EC) No 763/2008 of the European Parliament and of the Council (Text with EEA relevance)
- Commission Implementing Regulation (EU) 2017/543 of 22 March 2017 laying down rules for the application of Regulation (EC) No 763/2008 of the European Parliament and of the Council on population and housing censuses as regards the technical specifications of the topics and of their breakdowns (Text with EEA relevance)

Visualisation