43 patients undergoing total knee arthroplasty were randomized to three different tourniquet protocols to determine which was optimal for blood loss and fit-to-discharge outcomes. The three tourniquet applications were as follows: inflation just prior to incision and deflation after cement hardening (Group 1), inflation just prior to cementing and deflation after cement solidification (Group 2), and inflation just prior to skin incision and deflation after wound closure and compressive bandaging (Group 3). The results indicated that Group 1 had lowest blood loss and best outcome on the Timed Up-&-Go (TUG test; factor of discharge criteria).