68 patients undergoing total knee arthroplasty were randomized to either patellar subluxation (retraction) or patellar eversion during surgery. The purpose was to determine if subluxation of the patella led to any benefit over eversion in outcome in the short- and medium-term. Follow-up was conducted for 1 year postoperatively. The results demonstrated no significant differences in flexion, pain, and functional measures over 1 year postoperatively. Tibial overhang was significantly greater in the subluxation group, and two patellar tendon injuries were recorded in the subluxation group.