Cigarette smoking has been reported to protect women against pre-eclampsia. We conducted a cohort study of 493 women whose first antenatal visits were between September 1997 and April 1998 at Osaka Prefectural General Hospital, Japan. A self-administered questionnaire survey for lifestyles was carried out during pregnancy, and pregnancy outcome information was taken from medical record data. Of 493 subjects, 82 cases (16.6%) developed mild pre-eclampsia and 3 cases (0.6%) developed severe one. The prevalence of cigarette smokers decreased from 21.3% to 8.6% during early pregnancy. The incidence rate of pre-eclampsia among smokers was slightly greater than that among non-smokers (19.4% vs 17.1%), the incidence rate among cases exposed from household smoking was greater than that among no exposed cases (19.6% vs 14.3%), and the incidence rate among alcohol-drinkers was greater than that among non-drinkers (21.1% vs 15.1%). However, there were no statistically significant differences. Larger body mass index before pregnancy tended to be associated with the increased incidence rate of pre-eclampsia (p = 0.160). Pregnant women with smoking had a statistically higher frequency of household smoking exposure and having drinking alcohol. Household smoking exposure and drinking alcohol status adjusted hazard rate ratio was 1.1 for smokers (95% confidence interval 0.6-1.7) as compared with that for non-smokers. Our results did not support the proposition that cigarette smoking protected women against pre-eclampsia.