Our aim was to evaluate the optimal treatment strategy addressing cardiovascular risk in obese and nonobese patients with polycystic ovary syndrome (PCOS). We planned a prospectıve randomized clinical study. Normoandrogenemic and oligoamenorrheic women with PCOS and impaired glucose tolerance (n = 96) were enrolled in the study. Six months of treatment with metformin HCL or oral contraceptive pills (OCPs) were given to the patients. Group 1 were obese and receiving metformin. Group 2 were obese and receiving OCPs. Group 3 were nonobese and receiving metformin, and Group 4 were nonobese receiving OCPs. ADMA, homocysteine, high sensitive C-reactive protein (hs-CRP) and homeostasis model assessment estimate of insulin resistance (HOMA-IR) were investigated. ADMA, homocysteine, hs-CRP and HOMA-IR were similar in obese and nonobese groups before the treatment. After 6 months of treatment, a significant decrease was observed in ADMA, homocysteine and HOMA-IR levels in Groups 1 and 3. An increase in ADMA and hs-CRP levels was observed in Groups 2 and 4. In this study, metformin treatment leads to improvement in hormonal and metabolic parameters and decreases ADMA and homocysteine levels possibly independent of BMI. However, the use of oral contraceptives in obese and nonobese patients with PCOS with impaired glucose tolerance increases ADMA and hs-CRP levels and creates an increase in the metabolic risk.