OBJECTIVE To evaluate whether tourniquet release intraoperatively is better than postoperative release in reducing overall blood loss, duration of surgery, duration of tourniquet, length of hospital stay, wound related complications and transfusion requirement. METHODS Comparative study. METHODS Department of Surgery, The Aga Khan University Hospital, Karachi, from January 2004 to June 2007. METHODS One hundred and thirty patient files were reviewed retrospectively. Patients were divided into two groups. Group-A consisted of 65 patients with early deflation of tourniquet and group-B comprised of 65 patients with the release of tourniquet after applying compressive dressing. Total blood loss (determined by Gross method) and other study variables were noted as per objective and computed. RESULTS There were 22 males and 108 females with comparable BMI. All had undergone posterior stabilized cemented total knee replacement. Calculated blood loss was 1.208 L and 1.108 L in group-A and B respectively (p = 0.27). Significant increase in duration of surgery was noted in group-A patients. Four patients in group-B showed complication related to wound with 3 being minor and 1 requiring additional operation room visit. Mean length of hospital stay was 9 days. Transfusion frequency was higher in group-B despite comparable postoperative haemoglobin values. CONCLUSIONS Intraoperative tourniquet release does not reduce overall blood loss with no effect in conserving blood after total knee replacement, however, this group had relatively shorter hospital stay.