We performed gait analysis of 24 patients with bilateral knee osteoarthritis (OA) before, 1 year, and 2 years after bilateral total knee arthroplasty and compared these results to 12 normal controls. Cadence was significantly decreased in OA patients when compared with controls. Step and stride lengths were significantly different preoperatively (shorter) and postoperatively (P < .05), but postoperative results showed no significant difference when compared with normal controls ( P < .05). Step and stride times were significantly different ( P < .05) between the preoperative OA group and controls, but no difference was seen with the postoperative measurements when compared with controls. In the OA group, the mean vertical component of the ground reaction forces expressed as percent body weight was significantly lower when compared with controls and postoperative results. No correlation was found between the preoperative and postoperative clinical scores and objective data obtained front gait analysis. Our results demonstrate that gait analysis is not a useful tool in evaluating the success of total knee arthroplasty patient.