Tourniquet time and its relationship to blood loss and deep vein thrombosis were studied in 80 cases of unilateral total knee arthroplasty. These data were part of a prospective multicenter double-blind study on thromboprophylaxis for knee surgery. This study reports on the cases done at one institution in which the tourniquet was used in three different ways: group 1, no tourniquet use; group 2, limited tourniquet use for cementing time only; group 3, tourniquet use throughout the whole case. Blood loss was significantly related to tourniquet time (P = .0001). The incidence of deep vein thrombosis was not related to the tourniquet group (P = .9).