Background Few studies examine risk to offspring who experience both parental alcohol problems and parental separation, and still fewer consider gender of the affected parent. We examined interactive effects of maternal versus paternal alcohol problems and parental separation on timing of first alcoholic drink in daughters. Methods Data were drawn from a sample of 3539 European (or other) ancestry (EA) and 611 African ancestry (AA) female twins born between 1975 and 1985, median age 15 at first assessment. Cox proportional hazards regression models were estimated predicting age at first full drink from parental history of alcohol problems (mother only, father only, or both parents), parental separation during childhood, and the interaction of parental alcohol problems and parental separation. Cox models were estimated without and with adjustment for correlated risk factors, separately for EA and AA twins. Results For both EA and AA twins, a significant interaction between parental separation and mother only alcohol problems was observed, suggesting reduced risk of drinking associated with mother only alcohol problems in separated versus intact families. Conclusions Findings highlight parental separation as an important moderator of risk to children of mothers who have a history of problem drinking, with interactive effects observed consistently across racial group. To identify underlying processes, additional research is needed with more detailed characterization of separated families where mother only has a history of alcohol problems.