The relationship between optimal levels of glycemic control and perinatal outcome was assessed in a prospective study of 334 gestational diabetic women and 334 subjects matched for control of obesity, race, and parity. All women with gestational diabetes mellitus were instructed in the use of a memory-based reflectance meter. They were treated with the same metabolic goal according to a predetermined protocol. Three groups were identified on the basis of mean blood glucose level throughout pregnancy (low, less than or equal to 86 mg/dl; mid, 87 to 104 mg/dl; and high, greater than or equal to 105 mg/dl). The low group had a significantly higher incidence of small-for-gestational-age infants (20%). In contrast, the incidence of large-for-gestational-age infants was 21-fold higher in the mean blood glucose category than in the low mean blood glucose category (24% vs. 1.4%, p less than 0.0001). An overall incidence of 11% small-for-gestational-age and 12% large-for-gestational-age infants was calculated for the control group. A significantly higher incidence of small-for-gestational-age infants (20% vs. 11%, p less than 0.001) was found between the control and the low category. In the high mean blood glucose category an approximate twofold increase was found in the incidence of large-for-gestational-age infants when compared with the control group (p less than 0.03). No significant difference was found between the control and mean blood glucose categories (87 to 104 mg/dl). Our data suggest that a relationship exists between level of glycemic control and neonatal weight. This information is helpful in targeting the level of glycemic control while optimizing pregnancy outcome in gestational diabetes comparable to the general population.