To determine the effects of smoking on pre-eclampsia in an inner-city, predominantly African-American prenatal population. We performed a retrospective cohort study of all primigravidas who delivered singletons in our institution between 1980 and 1989. The study population consisted of 1862 smokers and 16 508 non-smokers. Univariate analysis and multiple logistic regression were used to identify relationships between smoking and adverse outcomes. The pre-eclampsia rate was significantly lower among smokers (11.3% vs. 13.0%, OR 0.85, 95% CI 0.73-0.99). Smokers with pre-eclampsia had higher rates of infants with low birth weight (OR 1.44, 95% CI 1.26-1.65) and very low birth weight (OR 1.85, 95% CI 1.55-2.20) and higher rates of placental abruption (OR 3.49, 95% CI 1.65-7.28) compared to non-smoking pre-eclamptics. This relationship persisted after correction for age, race, chronic hypertension and pre-eclampsia as confounding variables. Cigarette smoking was associated with a lower rate of pre-eclampsia among primigravidas independently of other maternal factors. Perinatal outcomes were significantly worsened among pre-eclamptics who smoked.