One hundred and sixty consecutive total knee arthroplasties were performed in 143 patients: 110 procedures, with a cemented kinematic-II prosthesis and fifty procedures, with a non-cemented porous-coated anatomic prosthesis. Each patient was evaluated before the operation and six weeks and three, six, twelve, and twenty-four months postoperatively. At a minimum twenty-four-month follow-up, the average Hospital for Special Surgery knee-rating score for the patients who had a cemented kinematic-II prosthesis was 9 points higher than the average score for the patients who had a non-cemented anatomic implant (88 points and 79 points). At the same follow-up period, the maximum flexion of the knees that had a cemented kinematic-II prosthesis was greater than that of the knees that had a non-cemented anatomic prosthesis (106 degrees and 97 degrees). In addition, the rate of reoperation for the patients who had a cemented kinematic-II replacement was 4 per cent, compared with 12 per cent for the patients who had a non-cemented anatomic prosthesis. On the basis of this prospective, non-randomized clinical review of unselected patients, we concluded that the results with the cemented kinematic-II prosthesis were superior to those with the non-cemented anatomic prosthesis at a minimum twenty-four month follow-up; however, these superior results may be related to the use of cement or to differences in the designs of the prostheses, the ages of the patients, or the postoperative management of the two groups of patients.