This study presents an obstetric protocol offering better management and prediction for normoglycemic insulin-dependent patients (White Class D4, F, R, or RF) who conceived after they were diagnosed as having vascular disease secondary to diabetes mellitus. Normoglycemia was accomplished during the pregestational phase, and conception occurred only after the glycosylated hemoglobin level was documented to be normal. Normoglycemia was maintained during pregnancy in the outpatient setting through the use of blood glucose monitoring performed by the patient. The obstetric protocol emphasized three additional areas of attention: (1) assessment of fetal growth by serial uterine fundal measurement and ultrasonography at gestational weeks 21 to 22; (2) assessment of fetal movement by patient-perceived fetal movements for 1 hour a week starting at week 35, increasing to 2 hr/day at week 37, and increasing to 3 hr/day from week 38 onward; and (3) cervical assessment at week 37 and preparation for vaginal delivery. Eight patients had a creatinine clearance of less than or equal to 80 ml/min prior to conception (mean = 66 +/- 6 ml/min). By 6 to 12 weeks' gestation all eight showed an increase in creatinine clearance (mean = 91 +/- 20, p less than 0.01). There was no change in the third trimester, and postpartum creatinine clearance was at antepartum levels. Proteinuria increased significantly by the end of the first trimester in all eight women and regressed post partum. Proteinuria (greater than 150 mg/24 hr) did not occur in the 14 women with normal antepartum creatinine clearance. Of 11 women with background retinopathy, six showed improvement in retinal status by fundus stereophotography whereas five showed no change. Of 11 women with proliferative retinopathy, five improved, five required laser therapy, and one remained in stable condition. Despite hemoglobin A1 levels in the normal gestational range (3% to 7.5%), there was a significant correlation of these levels with infant birth weights. None of the 22 infants died, and only one had any perinatal disease. Thus this protocol with its emphasis on fetal growth and size resulted in improvement in both maternal and infant outcome in pregnancies complicated by diabetes mellitus with vascular compromise.