AIM: The aim of the study is to compare the effects of metformin and insulin treatment for gestational diabetes mellitus (GDM) on vitamin B12 and homocysteine (Hcy) status. METHODS: Women with GDM, who met criteria for insulin treatment, were randomly assigned to metformin (n = 89) or insulin (n = 91) in the Adelaide cohort of the metformin in gestational diabetes (MiG) trial. Fasting serum total vitamin B12 (TB12 ), holotranscobalamin (HoloTC), a marker of functional B12 status and plasma Hcy concentrations were measured at 20-34 weeks (at randomization) and 36 weeks gestation, then at 6-8 weeks postpartum. RESULTS: Circulating TB12 , HoloTC and Hcy were similar in both treatment groups at each time point. Women who were taking dietary folate supplements at randomization had higher serum TB12 and HoloTC at randomization than those not taking folate. Overall, serum TB12 fell more between randomization and 36 weeks gestation in the metformin group than in the insulin group (metformin: -19.7 +/- 4.7 pmol/l, insulin: -6.4 +/- 3.6 pmol/l, p = 0.004). The decrease in serum TB12 during treatment was greater with increasing treatment duration in metformin-treated (p <0.001), but not in insulin-treated women. CONCLUSIONS: Total, but not bioavailable, vitamin B12 stores were depleted during pregnancy to a greater extent in metformin-treated than in insulin-treated women with GDM, but neither analyte differed between groups at any stage. This adds further evidence supporting metformin as a safe alternative treatment to insulin in GDM. Further investigation is needed to evaluate whether women treated with metformin for longer periods in pregnancy require additional B12 or other supplementation.