The results of twenty-two consecutive primary total knee replacements, performed an average of nine years (range, one to twenty-three years) after a patellectomy in twenty-two patients, were reviewed retrospectively. The average duration of follow-up was seven years (range, three to fifteen years). The average age of the patients at the time of the arthroplasty was sixty-seven years (range, thirty-six to eighty-nine years). The average Hospital for Special Surgery knee score was 46 points (range, 22 to 74 points) preoperatively and 76 points (range, 45 to 97 points) postoperatively. Thirteen patients had an excellent or good result and seven had a fair or poor result; in two patients, the operation was considered a failure. Except for one patient who had myasthenia gravis, all patients could climb stairs in reciprocal manner. Four patients lacked 5 to 20 degrees of active extension compared with passive extension. With respect to the over-all results, there was no significant difference among the four types of prostheses that were used (p = 0.2). The patients who had received an Insall-Burstein posterior stabilized prosthesis had better scores for pain and function than did the patients who had received a total condylar I prosthesis (p = 0.005 and 0.01, respectively). There was a direct correlation between the knee score and the number of years that had elapsed since the patellectomy. The longer the interval between the patellectomy and the total knee replacement, the higher the postoperative knee score (r = 0.78, p < 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)