PURPOSE: To compare clinical outcomes of infective and non-infective groups in revision total knee arthroplasty. MATERIAL AND METHODS: From December 1993 to December 2001, 29 cases of revision total knee arthroplasty were performed in 27 patients. The mean age at the time of surgery was 64.2 years (48~74). The average follow-up was 45.4 months (24~90). All cases were divided into two groups (fourteen noninfective and fifteen infective groups). The clinical results were evaluated according to range of motion, Hospital of Special Surgery score, Knee Society score and tibiofemoral angle. The causes of revision in noninfective group were aseptic loosening in eleven, instability in two, and femoral periprosthetic fracture in one. RESULTS: There was a significant improvement of range of motion, Hospital for Special Surgery Knee score and Knee Society score at the final follow-up compared with the pre-operative status. The above three evaluation criteria showed no significant difference between the two groups except higher further flexion in non-infective than infective group. Preoperative average tibiofemoral angle was varus 1.4degrees in non-infective group and valgus 2.6degrees in infective group (P<0.05). Postoperatively, the angles changed to be valgus 6.7degrees in non-infective group and valgus 6.8degrees in infective group(P>0.05). There were three complications in infective group, which were two reinfections, and one avulsion of osteotomized tibial tubercle. CONCLUSION: Preoperative planning, choice of proper implants, meticulous management of bony defect and soft tissue enabled successful results in infective group as well as non-infective group except less further flexion in infective group.