Oral anticoagulants are employed very frequently in the prophylaxis and treatment of several diseases. For their optimal effectiveness, many vigilance schedules have been proposed but none has proved to be 100% effective. The international normalization ratio (INR) can be a safer way to monitor oral anticoagulation, and our objective was to determine its clinical usefulness. A prothrombin time test (PT) was carried out by means of either a chromogenic or a coagulometric method, and an INR was obtained using the ratio of the PT patient/PT control elevated to an exponential given by the international sensitivity index (ISI) of our thromboplastin. Our objective was to maintain our patients in a therapeutical INR range between two and three. We present our experience with 77 patients and 810 results during an 18 months period. We observed 26 cases of hemorrhage and three of thrombosis. In all these cases, the INR was out of the desired therapeutical range. No deaths occurred in our patients. Our analysis showed a significative disagreement between the INR and the prothrombin time ratio (PTR) but a better correlation with hemorrhage and thrombosis was seen with the INR than with the PTR. Our experience supports the use of INR in the clinical vigilance of oral anticoagulation as a useful method.