Objective: this examined whether differences in perceived riskiness of alcohol consumption among a community sample of pregnant women. Further, this study examined the impact of prior experiences on risk perceptions, focusing on previous pregnancy experiences and on previous alcohol-related problems. Method: the hypothesized relationships among variables were tested simultaneously in a structural equation model. Subjects included 159 pregnant women, all of whom drank regularly before pregnancy recognition, who were recruited from prenatal clinics and through newspaper advertisements. Results: Perceived riskiness of drinking during pregnancy was lower among women who had previously given birth to a healthy child and among women with greater number of previous alcohol problems. Prior adverse pregnancy experience did not predict perceived risk. Perceived risk negatively predicted actual alcohol consumption during pregnancy, suggesting that previous healthy pregnancy experiences and alcohol problems increase drinking in pregnancy indirectly, through perceived risk. A direct positive effect from previous alcohol problems to drinking in pregnancy also was observed. Conclusions: Findings suggest that risk perceptions play a role in drinking behavior among pregnant women and help to illuminate the relationship between parity and alcohol consumption. Interventions designed to reduce drinking among pregnant women, which have generally relied on providing information, may be improved by considering the impact of previous experiences and addressing erroneous beliefs.