To determine whether women who continued to drink during pregnancy could be differentiated from women who discontinued alcohol use during their second trimester of pregnancy based on biological, social and behavioral data collected during a prenatal interview, 267 women receiving prenatal care at Grady Memorial Hospital, a large metropolitan hospital in Atlanta, were interviewed antepartum, assessing current drug and alcohol use as well as other demographic information. Postpartum interviews were conducted during the first 3 days following delivery to determine any changes in drug use or alcohol consumption that occurred after the first interview. Women who continued to drink throughout pregnancy and women who stopped drinking were similar on most demographic variables examined, including age, marital status, ethnic group, income, obstetrical complications risk score, amount of alcohol consumed per week and use of other drugs. Discriminant analysis was used to determine whether drinking-group membership could be predicted from self-reported drinking behaviors or biological and other demographic variables. The best predictors of drinking throughout pregnancy were the length of drinking history, reported tolerance to alcohol, a history of alcohol-related illness and drinking by siblings. In addition, women who continued to drink throughout pregnancy were more likely to report that they drank most often with other family members. Of the subjects who continued to drink, 81% were correctly classified based on this discriminant function. These findings suggest that women who continue to drink during pregnancy may be experiencing more chronic and severe alcohol-related problems than women who discontinue alcohol use and may thus be identified and targeted for intensive prevention effects.