Background. In 1918, it was observed that guanidine reduced blood sugar levelsin animals. Leter, a number of biguanidine derivatives of the general formula (R1, R2)-N-C(NH)-NH-C(NH)-N(R3,R4) were synthesized and pharmacologically tested. Only metformin remained for therapy of diabetes mellitus, while other biguanides wereabandoned because of their toxicity. Creatine, a final product of protein catabolism, was found to have a hypoglycemic effect in humans, but the results of its hypoglycemic activity were conflicting. Recent studies indicate that creatine, a naturally occurringguanidine compound, plays an important role in glucose homeostasis. The aim of this study is to demonstrate the antihyperglycemic effect of creatine and to compare the effect of creatine with the effect of metformin onglycemic control in patients with type 2 diabetes mellitus. Methods. A clinical study was performed on 30 patients with type 2 diabetes mellitus who were newly diagnosed and antidiabetic drugs naive. It was an openlabel, symmetrically randomised cross over study. On the first days visit, patients were not given therapy. In the first week of therapy, half of the patients received creatine (2x3g/day). Washout period was two days before cross over tometformin (2x500mg/day). The second half of the patients received metformin in the first week of the trial and after their two-day washout periods creatine. Blood glucose, insulin, C-peptid, creatine, lactate, binding on erythrocyte insulin receptors, fructosamine and Hb A1c were analysed during every visit at all control levels. Results. This study has already demonstrated that creatine is as effective as metformin in the efficiency of glycemic control. Its effect on blood glucose the concentration was statistically significant in all samples from patients treated with creatine when compared to patients without therapy (p