The purpose of this study was to determine whether an early third-trimester fetal abdominal circumference measurement can be used in patients with gestational diabetes to predict the presence or absence of macrosomia and labor dystocia at term. The predictive accuracy of a 30- to 33-week abdominal circumference measurement was tested, using the ninetieth percentile as the discriminant point. The study consisted of 201 patients with gestational diabetes who maintained weekly fasting glucose levels less than 100 mg/dl and 2-hour postprandial glucose levels less than 120 mg/dl with dietary management alone. The predictive accuracy of a 30- to 33-week fetal abdominal circumference measurement was 96.4% for ruling out macrosomia and 56.3% for predicting macrosomia. Patients with fetal abdominal circumference measurements greater than the ninetieth percentile at 30 to 33 weeks had a significantly increased incidence of cesarean section for failure to progress, shoulder dystocia, and birth trauma, whereas patients with abdominal circumference measurements less than or equal to the ninetieth percentile were at no greater risk than the general population. These results suggest that patients with non-insulin-dependent gestational diabetes with fetal abdominal circumference measurements less than or equal to the ninetieth percentile at 30 to 33 weeks are not at increased risk for macrosomia, cesarean section, or birth trauma at term, as long as their weekly glucose testing remains within normal limits. Efforts to decrease the incidence of macrosomia and its attendant risks should focus on those gestational diabetic patients whose fetal abdominal circumference greater than the ninetieth percentile at 30 to 33 weeks.