A retrospective matched pair study was designed to compare maternal perinatal factors such as abnormal pregnancy history and labor and delivery experience in families who subsequently were reported as physically abusive to one or more of their children as compared to non-abusive families. The study population consisted of the mothers of 532 children reported to the Baltimore (Maryland) Department of Social Services as physically abused during the years 1975-77. The comparison group was handmatched to the study group from State of Maryland birth certificates on the basis of the abused child's birth year and sex, maternal race, education and hospital of delivery. The study population was 67% black with a mean maternal education of 10.5 completed years. The abused children were 59% male with 48% less than 2 years of age. Results indicated that selected medical definitions of abnormal pregnancy, labor and delivery did not identify families at differential risk of maltreatment. However, mothers in maltreating families were younger, had shorter birth intervals, less prenatal care and were significantly more likely to have had a stillbirth or reported abortion or a prior child death. Study limitations are addressed as are suggestions for future research.