Patients starting oral anticoagulant therapy after heart valve replacement initially require a lower target international normalized ratio (INR) (2.0, range 1.5 to 2.6) because of a higher risk of bleeding until pericardial wires are removed. In a previous retrospective analysis, we observed a higher sensitivity to warfarin in these patients compared with nonsurgical patients. In a randomized clinical trial, we compared a fixed, lower dose of warfarin (2.5 mg) with the standard treatment consisting of a 5-mg loading dose, then adjusted to the target INR during the first 5 days of anticoagulation. INRs were measured daily, but the fixed dose was only modified on day 3 if the INR was <1.5 or >3.0. One hundred ninety-seven patients were considered eligible for the study. The 2 groups were well matched according to age, gender, body mass index, concomitant treatments, and type of valves implanted. The proportion of INRs >2.6 during the study period was 42.5% in the 5-mg group and 26.2% in the 2.5-mg group (p <0.05), and the proportion of INRs >3.0 on day 3 was 23.9% and 9.5% (p <0.05), respectively. In the 2.5-mg group, 35.7% of patients had an INR <1.5 on day 3 and had the dose increased (vs 3.5%, p <0.001); however, in the 5-mg group, 95.6% had the initial dose reduced, 49.6% had the dose withheld for at least 1 day, and the mean dose during the 5 days of study was 3.08 mg. Average time to achieve therapeutic range was higher in the 2.5-mg group (2.72 vs 1.98 days, p <0.0001), but the approach to the targeted INR was more regular, and the gap between target and mean INR on day 5 was smaller. There were no bleeding or thromboembolic complications in either group. Thus, a lower loading dose of warfarin in patients after heart valve replacement reduces excessive anticoagulation and offers a more regular achievement of the therapeutic target by reducing the number of dose adjustments. Daily monitoring of the INR is still recommended.