Sixty primary total knee arthroplasties in 43 Chinese patients were included into a prospective study. Twenty-six patients who had unilateral knee arthroplasty were randomized to receive continuous passive motion (CPM) or immobilization in the first week. The 2 groups of patients were comparable in demographic data and preoperative knee range of motion (ROM). In 17 patients who had 1-stage sequential bilateral arthroplasties, one side had CPM and the other side was immobilized. The active knee ROM was assessed regularly until 1 year after the operation. For all patients, the early active knee ROM in the CPM group was significantly better than the immobilization group. There was no difference after 7 days, however. For patients who had 1-stage bilateral total knee arthroplasties, the active knee ROM was significantly better on the CPM side until day 28. After 4 weeks, there was no difference between the CPM group and the immobilization group. Immobilization after total knee arthroplasty does not preclude good ROM.