This prospective investigation evaluated the reproducibility of the 100 gm oral glucose tolerance test. Sixty-four obstetric patients with greater than or equal to 135 mg/dl on the 50 gm oral glucose screening test were scheduled for the 100 gm test. All patients repeated the oral glucose tolerance test in 1 to 2 weeks. Both tests included a preparatory diet, and testing conditions were identical. There were no significant differences in the mean test values at each testing interval when the entire study population was considered. Patients were then divided into four groups according to the outcome of the two tests. Forty-eight of 64 (75%) had normal results at each testing period (group 1); 11 of 64 (17%) had initially normal results and abnormal results on retest (group 2); 3 of 64 (5%) had initially abnormal results and normal results on retest (group 3); 2 of 64 (3%) had abnormal results at both testing phases (group 4). There were no significant differences between oral glucose tolerance test results within groups 1 and 4. However, significant differences occurred within groups 2 and 3 between the two tests. Group 2 patients had a greater frequency of an abnormal 1-hour value on the test than group 1 patients (p = 0.001). Overall, the reproducibility of the oral glucose tolerance test was 78% (50 of 64). We recommend the oral glucose tolerance test be repeated when the 1-hour value is abnormal or when the fasting blood sugar, 1-hour, and 2-hour values are near the upper end of the normal range.