However, the risk can be avoided by giving up smoking, preferably before 35 years of age, experts said.
Researchers at the Oxford University and Japan investigated the impact of smoking on mortality in a large group of Japanese people living in Hiroshima or Nagasaki in 1950.
The findings, however, have nothing to do with radiation exposure from the nuclear bombs.
The Life Span Study (LSS) was initiated in 1950 to investigate the effects of radiation, tracking over 100,000 people.
However, most received minimal radiation exposure, and can therefore provide useful information about other risk factors.
Surveys carried out later obtained smoking information for 68,000 men and women, who have now been followed for an average of 23 years to relate smoking habits to survival.
Previous studies in Japan suggested smoking reduced life expectancy by only a few years compared with about ten years in Britain and the USA.
The younger a person was when they started smoking the higher the risk in later life.
Japanese born more recently (1920-45) usually started to smoke in early adult life, much as smokers in Britain and the USA.
These differences in smoking habits are reflected in the mortality patterns. Smokers born before 1920 lost just a few years.
In contrast, men born later (1920-45) who started to smoke before age 20 lost nearly a decade of life expectancy, and had more than double the death rate of lifelong non-smokers, suggesting that more than half of these smokers will eventually die from their habit.
Results on the women who had smoked since before age 20 were similar.
Nowadays, young Japanese smokers tend to smoke more cigarettes per day and to start at a younger age, so their risks will be higher, researchers said.
People who stopped smoking before age 35 avoided almost all the excess risk among continuing smokers, and even those who stopped around age 40 avoided most of it, they said.
The study was published in the British Medical journal (BMJ).
