The study aimed to determine the influence of glucose control during pregnancy on the incidence of macrosomia in the infants of mothers with insulin-dependent diabetes. The prevalence of macrosomia was determined in pregnancies of all such women attending the City Hospital Nottingham, between July 1987 and July 1993, where it is policy for diabetic patients to aim during pregnancy for preprandial capillary glucose levels of less than 6.0 mmol I-1, and postprandial glucose levels of less than 8.0 mmol I-1. Macrosomia was defined as a birthweight greater than 90th centile using a computer model which takes into account all the major determinants of birthweight. Twelve of 29 infants were macrosomic, which is greater than expected (p < 0.01). During pregnancy there was no difference in mean seven point glucose profiles between those mothers with normal weight babies and those with macrosomia, but fructosamine levels at booking were significantly higher in the latter 2.5 (1.9-2.9) vs 2.2 (1.2-3.0); p < 0.05. These data confirm other studies and suggest that the incidence of macrosomia may be reduced by tighter control of diabetes at conception and in the first trimester, but to a lesser extent during later pregnancy.