Based on information from two studies of sudden infant death syndrome (SIDS) from 1988-1991 in Tasmania, Australia, prospective and retrospective maternal responses to an identical set of questions were compared for 27 cases and 25 controls. There was good agreement on demographic factors, maternal obstetric history, parental smoking, and infant feeding practices. Reported changes in sleep habits were slightly greater for cases, and further work is needed to determine if this reflects recall bias or real changes during early infant life. Case mother reports regarding family history of disease and infant bedding were more discrepant, suggesting recall bias and supporting prospective collection of this information.