The question underneath
How long do smokers live?
The famous answer is "about ten years less". It is an average across habits that are nothing like each other, and the spread inside it is the useful part.
Written by Lee H. Last reviewed 31 August 2026.
The ten years comes from Jha et al., 2013, and it describes long-term smokers as a group. It is a real finding and it answers a question nobody actually has, because nobody smokes "as a group". Five a day for fifteen years and thirty a day for forty are both "smoking" and they are not the same exposure. So this page gives you the table instead.
By how much you smoke
A man who started at 20, smoked steadily, and is 60 now. Forty years of it. The only thing changing down the table is how many a day.
| A day | Years lost | Life expectancy |
|---|---|---|
| 0 (baseline) | 0 | 81.79 |
| 5 | 2.36 | 79.43 |
| 10 | 4.72 | 77.07 |
| 15 | 7.08 | 74.71 |
| 20 | 9.21 | 72.58 |
| 30 | 11.14 | 70.65 |
The range across that column is about nine years, which is nearly as wide as the famous ten. Somebody on five a day and somebody on thirty a day are both described by "smokers live ten years less" and neither of them is.
Why doubling the cigarettes does not double the damage
Read the table again and the numbers stop being a straight line partway down. Five to ten to fifteen a day steps up in even 2.36 year increments. Then 20 a day adds 2.13, down from 2.36, and 30 a day adds only 1.93 for ten extra cigarettes.
That is deliberate, and it is worth understanding. Below a certain total the model treats the cost as linear, straight from the measured per-cigarette rate. Above it, the curve bends and levels off toward a ceiling. Without that bend, a heavy sixty year habit would eventually produce a loss larger than a human lifetime, which is obviously wrong.
So the flattening is the model refusing to produce a number it cannot support, in the region where multiplying a per-cigarette rate stops making sense. The exact shape of that bend is a modeling choice, and it is written out on the methodology page so you can disagree with it.
The spread inside "about ten years" is nearly as wide as the ten years.
By how long you have been at it
Same man, same 15 a day, still smoking, at four different ages. The middle column is what the habit has cost by then; the right hand column is what he is forecast to reach if nothing changes.
| Age now | Years lost so far | Forecast |
|---|---|---|
| 40 | 3.54 | 75.05 |
| 50 | 5.31 | 74.59 |
| 60 | 7.08 | 74.71 |
| 70 | 8.77 | 75.89 |
The forecast column looks wrong at first glance and is not. It barely moves, and it even rises after 60. That is the life table doing its job: the baseline for a 70 year old is higher than for a 40 year old, because a 70 year old has already survived thirty more years of everything else that kills people. The loss column is the one that tells you what smoking did. The forecast column is the loss set against a baseline that is itself climbing.
This is the single most misread thing about life expectancy figures, and it is why "smokers die at 74" is not a sentence this site will write.
Men and women are not the same figure
The measured rate is about 17 minutes of life per cigarette for men and 22 for women (Jackson et al., 2025). Applied to the same 15 a day from 20 to 60:
| Man | Woman | |
|---|---|---|
| Baseline life expectancy at 60 | 81.79 | 84.73 |
| Years lost | 7.08 | 9.01 |
| Forecast | 74.71 | 75.72 |
She loses nearly two years more to the identical habit, and still forecasts about a year higher, because she started from three years further up. Both halves of that are true at once and neither is the whole story, which is the general problem with wanting one number.
The question worth asking instead
"How long do smokers live" is a question about a group you are only loosely a member of. The one that has an answer you can act on is how much of this comes back if I stop, and that depends mostly on your age.
Near all of it in your twenties. About 88.5% at 40. About 59% at 50. About 39% at 60. Those are the numbers with something to do attached to them, and there are separate pages on 40, 50 and 60.
The figures above are a motivational estimate built from population averages. They are not a prediction about you or about anybody, they do not diagnose, treat or prevent any condition, and this is not medical advice. No guarantee is made or offered about the accuracy of any figure or calculation on this site. Talk to a doctor or a quitline about stopping smoking. In the US, that is 1-800-QUIT-NOW.
Keep reading
Where to go next
Life expectancy calculator
Your own age, sex and habit.
Where the ten years comes from
The CDC statement, and the one paper underneath it.
What one cigarette costs
The measured rate, and why it cannot be multiplied forever.
Sources
- Jha, P. et al., "21st-Century Hazards of Smoking and Benefits of Cessation in the United States", New England Journal of Medicine, 2013.
- Jackson, S. E., Jarvis, M. J. and West, R., "The price of a cigarette: 20 minutes of life?", Addiction, 2025.
- US Social Security Administration, 2023 period life table, Social Security area population.