Infants of insulin-dependent diabetic mothers are considered to be at high risk for birth trauma, presumably due to macrosomia. With current management of diabetes in pregnancy, including strict glycemic control, the rate and the severity of macrosomia should be decreased. The frequent use of ultrasound to assess fetal growth and weight and the use of cesarean delivery in case of fetal macrosomia should further decrease the risk for birth trauma in these infants. We therefore undertook this study to test the null hypothesis that with current management, insulin-dependent diabetic mothers have a rate of birth trauma similar to that of infants of nondiabetic mothers (normal glucose challenge test at 28 weeks' gestation) matched for gestational age at birth, presence or absence of labor, delivery method (vaginal versus cesarean), and race. We studied 118 insulin-dependent diabetic mothers (White classes B-RT) and 354 control subjects (three matches for each insulin-dependent diabetic mother). The rate of birth trauma was 3.4% in insulin-dependent diabetic mothers, not significantly different from controls (2.5%). Logistic regression analysis in which birth trauma was the dependent variable and diabetes, race, presence or absence of labor, mode of delivery (vaginal versus cesarean), infant weight, and infant head circumference were independent variables revealed that only vaginal delivery was a significant risk factor for birth trauma in infants in both groups (p = 0.01). Most frequently observed birth traumas were brachial plexus injury, facial nerve injury, and cephalohematoma. Of the three infants with brachial plexus injury (insulin-dependent diabetic mothers, two; controls, one), two were delivered with use of midforceps.(ABSTRACT TRUNCATED AT 250 WORDS)