111 pregnant women with type-1 diabetes were cared for at the Karolinska Hospital from 1979 to 1986. As routine fetal monitoring, a non-stress test (NST) was performed twice weekly from the 35th or 36th week of gestation to delivery. If pregnancy complications occurred, an NST was still used for fetal monitoring, but more frequently. The median gestational age at delivery was 270 days. The mean maternal blood glucose during the third trimester was 6.0 mmol/l. In 88 of the totally 111 women the only indication for an NST was the patient's diabetes. In this routinely monitored group, 2/88 patients had abnormal NSTs and cesarean sections were performed. The neonatal outcome was good in both cases. Twenty-three had such complications as pre-eclampsia or IUGR, and in these cases the frequency of an NST was individualized. Four of these 23 had abnormal NSTs leading to cesarean sections. There were no signs of asphyxia among these four infants. Thus, in diabetic pregnancies with a well-regulated blood glucose, intervention due to abnormal fetal monitoring is more associated with acute pregnancy complications than the diabetes per se. The results of this study suggest that antenatal NSTs twice a week from week 35-36 is sufficient in well blood-glucose regulated type-1-diabetic women with a well-regulated blood glucose. If pregnancy complications occur, the NST, in an individualized frequency, seems to be a safe way of fetal monitoring.