The relationships between obstetric complications (OCs) and both family history of psychosis and season of birth were investigated among 70 demographically matched pairs of schizophrenic patients and control cases by using OC scores produced by three different OC scales. OCs were studied through blindly assessed hospital records. The particular OC scale had a great influence on the nature of the relationships observed between OCs and both family history and season of birth. Across the three scales, the findings varied from no relationship at all to completely opposite relationships between OCs and both of the other variables. More attention needs to be paid to OC methodological differences as a source of variation in study outcome, and attempts should be made to standardize methods for OC assessment across studies.