Over past years, there has been a world-wide increase in oral anticoagulant treatment (OAT). This study was aimed at evaluating the efficacy and safety of OAT managing in a real-practice situation. Nine hundred and three consecutive unselected patients referred for the control of OAT to the Anticoagulation Clinic of the University of Florence were studied. The total follow-up period was 1679 patient-years. The rate of total, major and fatal bleeding events was 5.0, 1.1 and 0.06 per 100 patient-years, respectively. In patients with a target International Normalized Ratio (INR) > or = 3, a significantly higher rate of bleeding (P = 0.02) with respect to patients with a target INR < 3 was observed. The rate of all thrombotic events was 3.8 per 100 patient-years. The rate of major and fatal thrombotic events were 2.4 and 0.4 per 100 patient-years, respectively. At INR >/= 4.5 the rate of bleeding was significantly higher (P = 0.005) than at lower INR. At INR < 2 the rate of all thrombotic events was significantly higher (P = 0.00001) with respect to more elevated intensities of anticoagulation. A low incidence of complications may be obtained even in elderly outpatients on OAT followed at an anticoagulation clinic.