One hundred consecutive primary total knee arthroplasties performed by a single surgeon were prospectively randomized into two groups to study the effect of tourniquet release for hemostasis on postoperative blood loss and transfusion requirements. The two groups were comparable. Variables such as antiinflammatory drug use; anesthetic; soft-tissue release; and component fixation were analyzed. There was no significant difference between the groups in terms of perioperative blood loss, decrease in hemoglobin or hematocrit level, transfusion need, or incidence of wound or thromboembolic complications. Tourniquet release for hemostasis is not an effective means of limiting postoperative blood loss or reducing transfusion need after primary total knee arthroplasty.