Accurate family histories of birth defects are critical for risk assessment and etiologic investigations. Typically, information about family history of birth defects is ascertained from interviews with birth mothers of index children; however, the quality of these interviews is rarely assessed. We evaluated family history information provided by case (n = 28) and control (n = 29) mothers who participated in a population-based, case-control study of orofacial clefts. Interview responses from mothers were compared to questionnaire reports collected by mail from first- and second-degree parental relatives. A total of 345 case and 380 control adult relatives completed questionnaires. These relatives also provided reports for 130 case and 169 control offspring. To examine the quality of birth defect reports, the sensitivity and specificity of birth mother responses were calculated. Sensitivity for presence (yes/no) of a birth defect was 31% for case mothers and 9% for control mothers. Specificity for case and control mothers was 98% and 97%, respectively. Interview responses from mothers who participate in family genealogy were more likely to be concordant with relative reports than were responses from mothers who do not participate in family genealogy. Case mother responses were more likely to be concordant than control mother responses. These results suggest that reliance on interview reports from birth mothers may lead to an underestimation of the occurrence of birth defects in relatives. Future investigations should explore methods to improve the quality of informant reports about family histories of birth defects. One alternative approach is discussed.