This study of 513 white cervical cancer patients and 490 white hospitalized controls considered, with adjustment for marital history and parity, the effects of smoking and dietary characteristics on the relative risk of developing cervical cancer. Although smoking was shown to be positively associated with risk, no dose-response relationship was evident. The frequency of consumption of cruciferous vegetables was positively associated with risk. Respondent reports of the frequency of prior ingestion of several food items were used in constructing crude indices of exposure to basic nutrients; the index of beta-carotene exposure was negatively associated with risk, whereas the index of fat consumption was positively associated with risk.