A prospective, controlled, randomized trial of continuous passive motion and immobilization after knee arthroplasty revealed that continuous passive motion significantly improved early and late flexion of the knee, reduced the duration of stay in the hospital, and did not increase the incidence of superficial infection or problems with wound-healing. Flexion of the knee beyond 40 degrees progressively diminished viability of the edges of the wound, particularly the lateral edge. On the basis of these results, a protocol for continuous passive motion was designed to minimize the detrimental effects on viability of the wound.