Premature delivery is common in pregnancies complicated by maternal diabetes. However, the outcome of very-low-birth-weight infants (VLBWI) born to mothers with diabetes is not known. Employing a matched double-cohort design, we investigated the influence of maternal diabetes on the outcome of VLBWI born in Winnipeg from 1988 to 1994. We compared mortality rates and early and late morbidity rates in VLBWI born to mothers with diabetes mellitus (DM) (cases, n = 43, 23 with gestational DM and 20 with pregestational DM) and without DM (controls, n = 539). Controls were matched for gestational age (GA), sex, and the year of birth. All subjects were enrolled in the Newborn Follow-Up Program. Relative risks and 95% confidence limits were calculated for each variable and Chi 2 analysis, Student t-test, and Mann-Whitney test were used as appropriate for analysis. Diabetes mellitus control was assessed by conventional criteria. There were no differences between cases and controls in mode of delivery, birth weight (mean +/- SD, 1,160 +/- 25 g vs 1,110 +/- 26 g), GA (29 +/- 2.8 wk vs 29 +/- 2.4 wk), smallness for gestational age (35% vs 30%), head circumference (26.5 +/- 1.9 vs 26.2 +/- 2.2 cm), length (38.8 +/- 2.8 vs 37.5 +/- 3.7 cm), Apgar score < 4 at 1 min (42% vs 40%) and < 7 at 5 min (37% vs 42%). Incidence of hyaline membrane disease (60% vs 71%), bronchopulmonary dysplasia (33% vs 31%), patent ductus arteriosus (30% vs 43%), necrotizing enterocolitis (12% vs 12%), sepsis (23% vs 25%), acute renal failure (9% vs 10%), intraventricular hemorrhage--all grades (74% vs 64%), retinopathy of prematurity--all stages (30% vs 26%), median days on ventilator (4 vs 4 days), and median days on supplemental oxygen (46 vs 42 days) were similar in both groups (p = NS, 95% confidence limits included 1 for all of these variables). There was no significant difference in mortality (21% vs 15%) or the incidence of major congenital anomalies. Weight, head circumference, and length at 6, 12, and 18 months were similar in both groups. There were no group differences in developmental quotients, prevalence of neurodevelopmental impairments, respiratory morbidity, or number of hospitalizations up to the last follow-up (18 months). Our data suggest that with contemporary perinatal care there is no significant increase in mortality rates or early and late morbidity rates between VLBWI born to mothers with DM and VLBWI of nondiabetic mothers. It seems that with reasonable diabetic control, prematurity rather than the diabetic state determines the neonatal outcome, and this knowledge can be useful in parental counselling.