The World’s Oldest Countries, and Who Is Aging Fastest

By the populationclock.org team · Published September 24, 2026 · All figures: UN World Population Prospects 2024, medium variant

The oldest countries right now

Median age is the cleanest single number for how old a country is: half the population is younger, half older. On the UN’s 2026 figures, Japan leads the countries of any real size at about 50.2 years, and it is the only one of the 128 entries with more than five million people to have crossed 50. Behind it come Italy at 48.6, Hong Kong at 47.9, Portugal at 47.2, Greece at 47.1 and Spain at 46.3.

A caveat before the list gets quoted. Across the whole dataset Japan ranks only 4th, because a handful of very small territories and micro-states post higher medians on populations of a few thousand. Those numbers are real but fragile, and they tell you about one town rather than about a society. Every comparison in this article is therefore drawn from the 128 countries and territories above five million people.

For scale: weight every national median by the people living under it and the world comes to roughly 31.7 years today, against about 22.5 in 1950. The UN builds its own global median differently, from the age distribution directly, so treat this as an indicator of direction rather than a headline figure. The direction is not in doubt. The same calculation gives about 42.8 by 2100.

It is worth noticing that the climb did not start where people assume. On the same weighting the world was actually getting younger for the first stretch of the record, bottoming out around 20.9 years in 1969 before turning. Birth rates were high enough almost everywhere that the base of the pyramid widened faster than the top lengthened. Global aging is a phenomenon of the last half century, not of the whole modern era, and it has been quick.

Who is aging fastest is a different list

Oldest and fastest-aging are not the same places, and confusing them is the most common error in coverage of this topic. Ranked by years of median age added between 2026 and 2050, the leaders in that set are Singapore, Hong Kong, China, Chile and Costa Rica. Singapore adds about 14 years of median age in that single stretch. China adds roughly 11, moving from 40.6 to 52.1.

The compression is easier to see as a journey between two fixed points. Italy took 73 years to travel from a median age of 30 to a median age of 50, crossing 30 in 1958 and 50 in 2031. Japan did it in 49. South Korea is projected to do it in 34, and China in 42. The end state is much the same; the time available to prepare for it is not.

Run that forward and Japan stops being the answer to the question in this headline. It is 1st today, 5th by 2050 and 14th by 2100, not because it gets any younger, but because its own median barely moves after mid-century, from 50.2 now to about 53.0, while the countries behind it keep climbing. China passes it to reach roughly 60.7 by 2100. Japan is not the extreme case of aging so much as the early one.

That is the substance of the East Asian story, which we cover at country level in Japan, China and South Korea. It is also why the fast list contains countries most people would not think to look for. Aging quickly is not a rich-country condition; it is what happens to any country whose birth rate falls steeply within a single generation.

What actually moves a median age

Two things, pulling in the same direction. Fewer births narrow the base of the age pyramid, which is by far the stronger effect and the fast-moving one. Longer lives lengthen the top, which matters less for the median than most people assume but compounds over decades. Weighted by population, life expectancy at birth worldwide has gone from roughly 49 years in 1950 to about 74 today on these figures, and the projection carries it to around 82 by 2100.

Japan illustrates both at once: life expectancy of about 85.1 years, among the highest anywhere, alongside a fertility rate near 1.23 births per woman. A population held up by long lives rather than by new arrivals ages quickly, then begins to shrink, which is what the countries that have already peaked have in common.

The youngest countries, and the gap

At the other end of the same set sits the Central African Republic, with a median age of about 14.6. The next several are also African: Somalia, Niger, Chad and DR Congo. Nigeria sits at 18.3, India at 29.2, the United States at 38.7. The distance between the oldest and the youngest entry in the set is about 36 years of median age, which is roughly a working lifetime. Two countries can be the same size, trade with each other and appear on the same charts while being a generation and a half apart in age structure.

Africa does age, substantially, and this is the part that gets missed. Its population-weighted median rises from about 19.9 years now to around 35.2 by 2100. But that end point is still below where Europe stands today, about 43.1, which is why the continent supplies so much of the century’s growth even as its own fertility falls. We follow that through in the African century.

The gap does not close

The intuitive guess is that everyone converges as fertility falls everywhere. On this projection they do not. The spread between the oldest and youngest entry in the set is about 36 years today and roughly 42 years in 2100, slightly wider rather than narrower. The whole distribution slides upward rather than bunching: 63 entries have a median age above 40 today, 117 do by 2050, and 173 by 2100. Above 50 the count goes from 5 to 70. Below 20 it goes from 38 to 0.

The population weighting tells the same story more bluntly. About 30% of people live in a country with a median age over 40 today; by 2100 it is roughly 56%. Japan has company coming: 10 entries in the set are above a median age of 50 by 2050 and 28 by 2100, led by Hong Kong, China and South Korea. The other end moves just as far. The youngest place anywhere in 2100 is projected to be Côte d'Ivoire, at about 30.5 years, a level 48 of the 50 European entries had already passed before 2000 and that 105 entries are still below today.

Two honest limits on all of this. A median age is a summary, not a description: it says nothing about how many people are working, how healthy the older half is, or what any of it costs. And these are projections, so the 2100 column rests on fertility and mortality assumptions that the same dataset publishes low and high variants for. The near-term figures are far firmer than the distant ones, for the plain reason that most of the people who will be middle-aged in 2050 have already been born. The mechanics behind these series are set out in our methodology, and the wider shrinking story in Europe’s quiet population decline.