OBJECTIVE: Studies of the effect of maternal smoking during pregnancy on development of brain tumors in the offspring generally have found no increase in risk but most have mainly relied on retrospective exposure assessment. We conducted a prospective study on a large birth cohort in Sweden. METHODS: Women giving birth during 1983-1997 were classified as smokers or non-smokers based on information ascertained at the first prenatal visit and recorded in the Swedish Birth Register. Follow-up of brain tumor incidence among offspring through 1997 was achieved by linkage with the Swedish Cancer Register. Hazard ratios were estimated using Cox proportional hazard regression, adjusting for demographic characteristics available in the Birth Register. RESULTS: Brain tumors (n=480) occurred at a rate of 4.5 cases per 100,000 person-years. Children of women who smoked during pregnancy had an increased incidence of brain tumors (hazard ratio = 1.24; 95% confidence interval: 1.01-1.53). The increase in risk was similar for benign and malignant tumors, and was most apparent for astrocytoma. The effect of smoking on the occurrence of brain tumors was seen most strongly among 2-4 year-old children. CONCLUSIONS: These results support a role for maternal smoking during pregnancy in the etiology of childhood brain tumors. Our findings should be confirmed in other prospective studies.