Proprioception was measured in 2 groups of patients after successful knee arthroplasty. Twenty-eight patients had total knee arthroplasty and their results were compared with an age matched group of 10 subjects who had undergone unicondylar knee arthroplasty. The threshold to detection of passive motion was quantified as a measure of proprioception. The degree of preoperative arthritis was objectively classified according to Resnick. The anterior cruciate ligament and posterior cruciate ligament were present and preserved in all the patients who had undergone unicondylar knee arthroplasty. The anterior cruciate ligament was sacrificed and posterior cruciate ligament retained in 15 of the patients who had total knee arthroplasty and the anterior cruciate ligament and posterior cruciate ligament were sacrificed in 13 of the patients who had total knee arthroplasties. There was no difference in threshold to detection of passive motion among any of the 3 groups. Maintaining the anterior cruciate ligament and posterior cruciate ligament did not impart improved proprioception in unicondylar knee arthroplasty nor did maintaining the posterior cruciate ligament impart improved proprioception in total knee arthroplasty.