Twenty-six total knee arthroplasties (TKAs) were evaluated in 22 patients who had had a patellectomy. Fourteen knees (12 patients) had a primary TKA, and 12 patients had a revision TKA. Two patients in the revision group, whose prostheses failed, were from the primary TKA group. The mean follow-up time was 8.5 years in the primary TKA group and 7.6 years in the revision TKA group. A group of 14 control knees with patellae was randomly generated but matched for prosthesis, diagnosis, surgeon, age, and time of surgery. This group was similarly evaluated with an average follow-up time of 6.9 years. The primary TKA group had seven knees that were rated as good or excellent, two as fair, and three as poor. The control group had a significantly higher average rating than the primary TKA group. In this group, there were 12 good or excellent knees, three fair, and none poor. Postoperative pain, flexion contracture, extension lag, and range of motion all contributed significant information to the final score, whereas other variables (walking, function, strength, and instability) did not contribute any additional information. Although higher overall scores may have been expected if the patients had patellae, the results during the follow-up examination were satisfactory and justified TKA in these patients. In general, however, patients without patellae may be at a higher risk for failure of the prosthesis, as seen in five patients having primary TKA and another ten patients with failed TKA requiring revision.