Patients with gestational diabetes were divided into two groups according to the results of three-hour oral glucose tolerance tests. Those with fasting euglycemia (serum glucose 95 mg/dL or lower) on oral glucose tolerance test (class A) were treated with diet alone, whereas those with fasting hyperglycemia on oral glucose tolerance test (class A/B) were treated with both diet and insulin (15 U neutral protamine Hagedorn insulin before breakfast). The frequency of macrosomia (birth weight more than 4000 g) among class A/B gestational diabetics was 16.2%, which was significantly greater than the 5.6% incidence in class A diabetics and the 6.4% incidence in controls. After controlling for potential confounding risk factors, it was determined that class A diabetics had a frequency of macrosomia no different from that of nondiabetics. Nonobese gestational diabetics with fasting hyperglycemia (class A/Bs), who were treated with diet and prophylactic insulin, also had a frequency of macrosomia no different from that of nondiabetics or class A diabetics. However, the diet and insulin regimen did not prevent excess macrosomia in class A/B diabetics who were obese.