To determine whether metformin, when given to women with polycystic ovary syndrome (PCOS), promotes folliculogenesis by prompting a drop in free sex steroids resulting in a compensatory follicle stimulating hormone (FSH) rise, we conducted a randomized, double-blind, placebo-controlled crossover clinical trial. Eight mid-reproductive age PCOS participants with mean obese body mass index (BMI) and normal glucose tolerance received 8 weeks of metformin, given in a step-up fashion to a maximum dose of 2000 mg daily or placebo with daily urine sampling, 4-6 weeks washout, and crossover to the remaining arm for 8 weeks. To confirm the effects of metformin on glucose and other metabolic markers, a hyperinsulinemic, euglycemic 3-dose clamp (physiologic: 30 mU/m(2) per minute, high: 400 mU/m(2) per minute) followed each treatment. Urinary FSH, luteinizing hormone (LH), or pregnanediol glucuronide (Pdg) did not differ by treatment. Glucose disposal, endogenous glucose production, BMI, ovulation rates, serum sex steroids, free fatty acids, and lipids did not significantly differ by treatment, despite good evidence for compliance with the protocol. During the clamp, high-dose insulin administration was associated with an acute drop in serum LH. We conclude that short-term, high-dose metformin exerts minimal effects on both metabolic markers and reproductive hormones in a small sample of overall morbidly obese women.