A consecutive series of revision total knee arthroplasties done at two centers was evaluated for the presence of heterotopic ossification on radiographs taken before and after revision using the classification system of Harwin et al. Knee Society scores were obtained preoperatively and at annual intervals postoperatively. The patients' demographics and clinical scores were correlated with the incidence and grade of heterotopic ossification. Minimum 2-year followup was obtained in 135 of 151 patients who had revision total knee arthroplasty during this period (89%). The incidence of heterotopic ossification before revision surgery was 23%, which increased to 56% at most recent followup (mean, 30 months; range, 24-48 months). The only risk factor identified for the development of heterotopic ossification was the presence of infection (76%), which was significantly higher than the 47% incidence of heterotopic ossification in patients who did not have an infection. The average postoperative Knee Society score was lower in patients with heterotopic ossification compared with patients without heterotopic ossification (129 points versus 148 points). Patients with heterotopic ossification had significantly lower functional scores particularly on stair climbing but did not have a significantly decreased range of motion. Parameters not associated with subsequent development of heterotopic ossification included gender (males), patient size (body mass index), surgical time, operative approach, or number of prior knee procedures.