The purpose of this study was to compare unicompartmental knee arthroplasty (UKA) with total knee arthroplasty (TKA) and more specifically to evaluate the role of the patella in patient preference between UKA and TKA. A group of 23 patients were chosen, each with a UKA in one knee and a TKA in the opposite knee. As a subset of the group, 13 patients were compared who had not had patellar resurfacing on their TKA side (Group A) versus ten patients who had patellar resurfacing (Group B). Each patient had a UKA and TKA performed during the same hospitalization. Each patient's resurfacing was performed by the same surgical team. Moreover, inpatient care and physical therapy for each patient's respective UKA and TKA were the same. Patient evaluation consisted of chart review, joint registry data, and telephone interviews that focused on patient preference regarding pain, stability, "feel," and ability to climb stairs. The 23 patients studied had an average follow-up period of 81 months (range, 38-153 months). There were 14 men and ten women with an average age of 67 years. Preoperative diagnosis was osteoarthritis in 22 patients and rheumatoid arthritis in one patient. Range of motion (ROM) improved from a preoperative mean of 106 degrees to 123 degrees postoperatively on the UKA side. Mean ROM for the Group A TKAs improved from 104 degrees to 109 degrees, whereas the Group B TKAs remained unchanged at 113 degrees. For patients surveyed in Group A, 31% stated that their UKA knee was their better knee overall, 15% stated that their TKA knee was their better knee overall, and 54% could find no difference.(ABSTRACT TRUNCATED AT 250 WORDS)