In this study we sought to test the hypothesis that treatment of women with one abnormal oral glucose tolerance test value will result in reduction of adverse outcome. One hundred twenty-six women with one abnormal oral glucose tolerance test value and 146 women in the control group (normal oral glucose tolerance test values) participated in a prospective study during the third trimester of pregnancy. The subjects with one abnormal test result were randomized into treated (group 1) and untreated groups (group II). Group 1 subjects were treated with a strict diabetic protocol to maintain tight glycemic control by means of diet and insulin therapy. Group 2 subjects tested their capillary blood glucose for a baseline period. The study revealed that the level of glycemic control was similar before initiation of therapy (mean capillary blood glucose 118 +/- 14 vs. 119 +/- 15 mg/dl, p = NS) for groups 1 and 2, respectively. There was a significant difference in mean capillary blood glucose (95 +/- 10 vs. 119 +/- 15 mg/dl, p less than 0.0001), preprandial, and postprandial determinations between the treated and untreated groups. The overall incidence of neonatal metabolic complications (4% vs. 14%, p less than 0.05) and large infants (6% vs. 24%, p less than 0.03) was significantly lower in the treated group. Comparison between the control (normal oral glucose tolerance test) and the untreated groups showed a significantly higher incidence of large infants and metabolic complications. No difference was found between the normal and treated groups. Thus we conclude that treatment of individuals with one abnormal oral glucose tolerance test value will result in significant reduction in adverse outcome in pregnancy.