Women with recent gestational diabetes mellitus were randomly assigned to one of two low-dose oral contraceptives to evaluate the effect of low-dose oral contraceptives on carbohydrate and lipid metabolism. A cohort of similar women requesting a non-oral-contraceptive method served as controls. The two oral contraceptives studied were ethinyl estradiol (0.035 mg)-norethindrone (0.40 mg) and ethinyl estradiol (0.030 to 0.040 mg)-levonorgestrel (0.050 to 0.125 mg). A 75 gm, 2-hour oral glucose tolerance test and a fasting lipid profile (total cholesterol, triglyceride, high- and low-density lipoprotein cholesterols) were performed at entry, after 3 months, and after 6 to 13 months of treatment. The prevalence of diabetes at 6 to 13 months (27/156 patients) was not significantly different between groups (non-oral-contraceptive group, 17%; ethinyl estradiol-norethindrone, 15%; ethinyl estradiol-levonorgestrel, 20%). When examined by prior gestational diabetes mellitus class, diabetes mellitus was present in 7% of prior class A1 and 29% of women with prior class A2 disease (p less than 0.001). Mean cholesterol and low-density lipoprotein cholesterol levels were significantly improved in all three groups at 3 months and at 6 to 13 months, whereas triglycerides remained unchanged. There were no differences in cholesterol, low-density lipoprotein cholesterol, or triglycerides levels between the groups. After 6 to 13 months, there was a significant increase in high-density lipoprotein cholesterol in the ethinyl estradiol-norethindrone group compared with the ethinyl estradiol-levonorgestrel and non-oral-contraceptive groups.