Abstract: In surgical practice, hemostasis is used to minimize postoperative bleeding in total knee arthroplasty. We performed a prospective randomized study to determine the influence of tourniquet use and the timing of its release on amount of bleeding. Eighty four patients (96 knees) were scheduled for total knee arthroplasty and randomly divided into three groups. Posterior cruciate retaining bicompartmental total knee prostheses were used in all. In group I, no tourniquet was used. In group II, a tourniquet was used and was deflated for hemostasis once all components had been inserted. In group III, the tourniquet was deflated after wound closure and application of a compressive dressing. Mean blood drainage, mean volume of blood transfusion, hemoglobin (Hb) and hematocrit (Hct) values and operative time were compared between the three groups. Mean blood drainage was 810 mL (300-1300) in groups I 720 mL (240-1200) in group II and 705 mL (250-1150) in group III (p = 0.062). The Hb and Hct values, tourniquet time (for groups II and III) and volume of blood transfusion were similar. The operative time was significantly longer for first group (p = 0.012). Using tourniquet and its intraoperative release with hemostasis, does not reduce blood loss in total knee arthroplasty, but using tourniquet reduces operation time significantly.