Four Numbers Called 'Average Age of Death'

If you have searched for the average age of death for a UK male and come away with three or four different numbers, you were not doing it wrong. There is no single "average age of death". There are at least four figures that all get called that, they are built in different ways, and they answer different questions. A man of forty who has just watched his father die at sixty-one is usually asking one very specific question, and only one of the four numbers actually answers it.

The confusion matters because people use the wrong number to frighten or reassure themselves. Someone reads that average life expectancy for men is in the late seventies, subtracts their current age, and quietly plans around the difference. That sum is almost always wrong, because the figure they started with was not the figure they needed.

So here is the map. The four numbers are the mean age at death, the median age at death, period life expectancy and cohort life expectancy. This article explains what each one measures, where each one comes from, and which one you want if you are measuring yourself against a parent who died young.

The short answer: four numbers, not one

Ask the Office for National Statistics how long a man lives and you will not get a single reply, because the honest answer depends on what you mean by the question. Do you mean the age at which men are currently dying? The age a newborn boy can expect to reach on today's death rates? The age that same boy will probably reach once you allow for medicine improving over his lifetime? These are genuinely different questions, and the ONS publishes different numbers for them.

The four figures line up like this. The mean age at death is the simple average age of everyone who actually died in a given year. The median age at death is the age at which half of a group have died and half are still living. Period life expectancy is a projection built by taking this year's death rates at every age and applying them to a hypothetical person for their whole life. Cohort life expectancy does the same but lets the death rates fall over time, as they generally have.

None of them is more "true" than the others. They are answers to four questions, and the mistake is to grab whichever one appears first and treat it as the answer to yours.

The mean and median age at death

The mean age at death is the one that sounds most like plain arithmetic: take everyone who died last year, add up their ages, divide by the number of people. It is a real figure and the ONS reports it. But as a guide to your own future it is close to useless, and it is worth understanding exactly why, because the reason is not obvious.

The mean is pulled down by every death at every age. A baby who dies in its first week, a cyclist killed at nineteen, a woman who dies of breast cancer at forty-four — all of them go into the average and all of them drag it downwards. So the mean age at death of the population is lower than the age a healthy adult should plan around, because you have already survived the years that pull the number down.

This is the single most common error in the whole subject. A man of fifty reads a mean age at death somewhere in the late seventies, does the subtraction, and concludes he has under thirty years left "on average". But that average included people who never reached fifty at all, and he is not one of them. Having got to fifty in reasonable health, his own expected age at death is higher than the population mean, not equal to it.

Source: Office for National Statistics, National life tables, UK and mortality statistics, published under the Open Government Licence v3.0.

The median, and why it usually sits higher

The median age at death is the tipping point: line everyone up by the age they died and the median is the age of the person in the middle. Because deaths cluster heavily in old age, with a long tail of earlier deaths pulling the mean down, the median age at death tends to sit a little above the mean. It is a slightly more intuitive "typical" age than the mean for that reason, though it carries the same health warning — it describes the ages people are dying at now, not the age you specifically will reach.

The gap between mean and median is a clue to the shape of the data rather than a fact about you. When a distribution has a long low tail — a minority dying young — the mean is dragged below the median. Deaths in the UK do exactly that. So if you see two "average age of death" figures a couple of years apart, one is probably the mean and one the median, and neither is the number a middle-aged person should be planning around.

Period life expectancy: today's rates, frozen in place

Period life expectancy is the figure most people have actually seen, because it is the headline the ONS leads with. In the most recent national life tables, period life expectancy at birth is a little under 79 years for men and a little under 83 for women. But the way it is built is widely misunderstood, and the misunderstanding changes what the number means.

Period life expectancy takes the death rates observed in one short window — the ONS uses three-year windows — and imagines a person living their entire life under those rates, unchanged, from birth to death. It is a snapshot of current mortality expressed as a lifespan. It deliberately does not assume medicine will improve, because its job is to summarise today's mortality in one comparable number, not to forecast any real person's life.

That design has a consequence people miss. Because it assumes no future improvement, period life expectancy tends to understate how long today's babies will actually live, since death rates have generally fallen over the long run. It is the right tool for comparing this year with last year, or one region with another, precisely because it holds everything else still. It is the wrong tool for guessing your own age at death, because your life will not be lived under frozen rates.

Source: Office for National Statistics, National life tables, UK: 1980–1982 to 2022–2024, published under the Open Government Licence v3.0.

Cohort life expectancy: allowing for improvement

Cohort life expectancy is the fourth number and the one closest to a genuine forecast for a real person. Instead of freezing this year's death rates, it follows an actual birth cohort forward and applies the death rates each age group is projected to face in the year they reach it, allowing for expected improvements in mortality along the way.

Because it lets death rates fall over a lifetime, cohort life expectancy comes out several years higher than the period figure for the same person. It is the number a pension scheme or an insurer leans on when it needs to know how long people will really live, not how long they would live if nothing ever got better. The ONS publishes it separately, as a projection, and flags the uncertainty around it — improvements slowed sharply during the 2010s, so the projections are not the confident upward march they once were.

If your instinct is that you will probably live longer than the headline life expectancy suggests, cohort life expectancy is the number that agrees with you. It is also the reason a healthy person planning to die at "the average" is usually planning to die too early.

Which number you actually want

Now the question that brings most people here. You are perhaps forty, perhaps fifty, and a parent died young. You want to know what age you can reasonably expect to reach. None of the four numbers at birth answers that, because all of them either describe a population you are no longer statistically part of or a newborn you no longer are.

The figure you want is life expectancy at your current age. The ONS life tables give it for every age, and it is always higher than the number quoted at birth, because it only counts people who, like you, have already survived to that age. Reaching forty or fifty in reasonable health means the deaths that dragged the birth figure down no longer apply to you. Your remaining life expectancy is measured from where you actually stand.

This is the honest reframing for anyone benchmarking themselves against a dead parent. A father dying at sixty-one is a real signal worth taking seriously, and it is worth telling your GP about, but it does not reset your own life expectancy to sixty-one. Start from life expectancy at your age, then think separately about whether there is a specific inherited reason to adjust it. We work through that layering in detail for people whose parents both died before they were 60, where the baseline probability comes first and the family history is added on top rather than replacing it.

The distinction also matters when the death in the family was sudden and unexplained rather than the end of a long illness, because the response there is clinical screening rather than actuarial arithmetic. We set out that pathway separately for people dealing with a brother's sudden death that nobody can explain.

Why the number at birth misleads almost everyone

It is worth being explicit about the trap, because it is so widespread. Life expectancy "at birth" is the default headline, and it includes every death from the first day of life onwards. Infant and childhood deaths are rare now but they still weigh on the figure, and so does every death in early and middle adulthood. All of that is baked into the birth number.

The moment you have survived past those ages, the birth number stops describing you. This is not a statistical technicality; it changes the answer by years. A woman of sixty has an expected age at death meaningfully higher than the life expectancy at birth for women, because she has already cleared six decades of risk. The older you get, the more the gap widens, which is why the very old routinely outlive the "average" they were quoted at birth by a wide margin.

For most working-age readers the practical lesson is simple. Ignore the birth figure for planning. Find life expectancy at your age, treat it as a population midpoint rather than a personal verdict, and remember that a midpoint means half of comparable people live longer still.

What this means when you are weighing up cover

These four numbers get misused in two opposite directions, and both cause trouble. Someone who fixates on the low mean age of death talks themselves into believing time is short and rushes decisions. Someone who reads the higher cohort figure decides death is a distant abstraction and puts off arranging anything at all. Neither reaction is what the numbers actually support.

The useful truth sits in between. Death in any single working-age year is genuinely uncommon, which is why the mean is not close, but the cumulative chance across the decades that a mortgage or a young family spans is far from trivial. That combination — low each year, meaningful over a term — is exactly the shape that a fixed-term life policy is built for. It is also why the amount matters as much as the fact of cover: work through whether four times salary in death in service is enough for your family before assuming an employer's benefit has it handled, because it usually falls short of a mortgage plus dependants.

The averages are just as easy to misread at the other end of life. Living to a good age does not mean money simply passes cleanly to your children; a death past seventy-five changes how a pension is taxed in the hands of whoever inherits it, which we cover in what your children are left with after tax if you die past 75. If you have decided the sensible move is to price cover against your own age and circumstances rather than a population average, you can get life insurance quotes and have an FCA-authorised broker work from your real figures. The rest of this subject, including how family history sits alongside the raw odds, is gathered in our guide to family history, inherited risk and the odds.

Scotland, Northern Ireland and where you live

Life expectancy is not uniform across the UK, and the differences are large enough to matter more than the mean-versus-median distinction. Figures for Scotland run below those for England, and Northern Ireland and Wales sit on their own paths, with the gap driven by a mix of deprivation, industrial history and health. Within any nation, the spread between the most and least deprived areas is wider still than the spread between the four nations.

The ONS, together with National Records of Scotland and the Northern Ireland Statistics and Research Agency, publishes life expectancy broken down by nation and by local area. If you are trying to place yourself on the scale, the national headline is a blunt instrument; the figure for people of your age, sex and area is closer to the mark, though it is still a population average and still not a prediction about you. You can read the underlying tables at the Office for National Statistics.

Frequently asked questions

What is the average age of death for a man in the UK?

It depends which average you mean. The mean age at death of people who died last year is pulled down by deaths at every age, so it undershoots what a healthy adult can expect. Period life expectancy at birth assumes today's death rates never improve. Cohort projections allow for future improvement and give a noticeably higher figure. All three describe populations rather than any individual.

Which figure should I use to compare myself with a parent who died young?

Use life expectancy at your current age, not at birth. The figure quoted at birth includes everyone who died in infancy and childhood, so having already reached forty or fifty means your own expected age at death is higher than the headline number implies. A parent's early death is worth mentioning to your GP, but it does not reset your own life expectancy to their age.

Is life expectancy still rising in the UK?

Improvements slowed markedly during the 2010s, and the pandemic period disrupted the series further. ONS national life tables are published as three-year averages partly for that reason. The long trend remains upward across decades, but the rapid year-on-year gains of the late twentieth century have not continued, which is why recent cohort projections are more cautious than earlier ones.