This study compares two methods of determining parental alcoholism: (1) a single question asking whether one of the subject's parents had alcohol problems; (2) the Family History Research Diagnostic Criteria (FH-RDC), which require at least one alcohol-related problem in addition to parental problem drinking. We sought first to determine the level of agreement between the single-question approach and the full FH-RDC for parental alcoholism, and then to gauge how much bias will be introduced by using the single question rather than the FH-RDC-method in assessing the risk for psychiatric disorders. Data were taken from the National Comorbidity Survey (NCS), a nationwide survey of the noninstitutionalized US civilian population aged 15 to 54. Parental alcoholism was assessed both by the single question and by the full FH-RDC. Psychiatric disorders were assessed with the Composite International Diagnostic Interview (CIDI). Good agreement was found between the single-question approach and the FH-RDC (kappa = 0.83). The single-question method introduces only a small downward bias when it comes to assessing the risk for psychiatric disorders in children of alcoholic probands. This underestimation does not reach statistical significance. When a questionnaire needs to be shortened, a single question on parental problem drinking is a good alternative to checking all individual FH-RDC items.