A comprehensive population study of women in Goteborg, Sweden was carried out in 1968--1969. Altogether 1462 women in five age strata were studied. The age stratum 50 comprised about the same number of postmenopausal women (n = 161) and women who still menstruated (n = 168). A comparison could, therefore, be made between still menstruating and postmenopausal women of almost identical age. In postmenopausal 50-yr-old women 50% were smokers compared to 26% in 50-yr-old women who still menstruated. The difference was highly significant. Almost all the postmenopausal women had been smokers for many years, which means that they did not start smoking as a consequence of the menopause. It thus seemed that an earlier menopause was a consequence of smoking. The difference could not or only to a small extent be explained by other factors such as differences in body weight between smokers and non-smokers.