The effect of maternal diabetes on the risk of congenital malformations was investigated in 23,695 pregnancies of white mothers, inclucing 339 patients of the Joslin Clinic, and in 24,742 pregnancies of Negro mothers, drawn from the prospective Collaborative Perinatal Project. Of these, 372 mothers had gestational diabetes and 567 had overt diabetes (before and during pregnancy). Pregnancy outcomes studied included stillbirths and live births. Among core women (excluding the Joslin Clinic cases), the frequencies of diabetic pregnancies were 1.31% and 1.18% for white and Negro mothers, respectively; in both groups, approximately two thirds of these pregnancies had gestational diabetes. There was no increase in malformation risk in the pregnancies of mothers with gestational diabetes over that of nondiabetic mothers in either racial group. However, the risk of malformation for white mothers with overt diabetes was double that of nondiabetic mothers for both major and minor categories of malformations. The incidences of major and minor types of malformations in the pregnancies of white mothers with overt diabetes were 17.94% and 10.94%, respectively, compared to the corresponding incidences of 8.34% and 6.25% for the white nondiabetic group. In Negro mothers with overt diabetes, a smaller increase of risk was seen only in major malformations; the incidences were 13.64% and 8.45% for the diabetic and nondiabetic groups, respectively. The increased risks for malformations were distributed generally throughout the organ systems. Multiple malformations occurred more frequently in the overt diabetic than in the nondiabetic group, suggesting that maternal diabetes must act adversely an an early stage of fetal development. Two cases with the caudal regression syndrome were observed in children of diabetic mothers, whereas none was found among births from nondiabetic mothers. Insulin (or analog) therapy of diabetes neither decreased nor increased the risk of malformation in the fetus. However, duration of diabetes had a significant effect on the malformation risk: the longer the mother had the disease, the higher was the incidence of malformations in the fetus. Paternal diabetes did not contribute to increase in risk. These observations suggest that maternal diabetes per se, through its adverse effects on maternal metabolism, is the responsible factor for the increase of malformations in the offspring.