One hundred seventy-six patients with osteoarthritis of the knee were randomized prospectively into two groups. In both groups the posterior cruciate ligament was released from its femoral attachment. In one group a posterior stabilized tibial component was used whereas in the other group a deep-dish tibial polyethylene component was inserted (Genesis II). The surgical and perioperative technique was identical in both groups and all the implants were cemented to their respective bones. Patients began range of motion exercises within the first few hours after surgery and were allowed weightbearing to tolerance beginning on the first postoperative day. At followup there was no statistical difference in the mean range of flexion (approximately 116 degrees), ability to ascend and descend stairs in a bipedal manner (80%), pain scores, knee scores (94 points), stability, or the lack of anterior knee pain. Postoperative implant alignment in the sagittal and coronal planes and on Merchant skyline views was excellent in both groups. There was only one lateral release required and that was in one patient who received a deep-dish component. Using deep-dish implant obviates the need to resect intercondylar femoral bone, decreasing the potential for fracture and maximizing bone volume should revision be necessary in the future.