7We previously reported the minimum eight-year follow-up results of cruciate-retaining total knee arthroplasty in a consecutive series of seventy-two knees in patients with rheumatoid arthritis. In the present study, we evaluated the longer-term outcomes after twenty to twenty-five years of follow-up. Since the publication of our original study, ten knees have been revised: three because of periprosthetic fracture, three because of infection, two because of patellofemoral failure, and two because of posterior instability. The rate of implant survival at twenty years after surgery was 69% (95% confidence interval [CI], 56% to 79%) with revision for any reason as the end point, 81% (95% CI, 69% to 89%) with femoral or tibial component revision for any reason as the end point, and 93% (95% CI, 83% to 97%) with posterior instability as the end point. These long-term results demonstrate that posterior cruciate ligament insufficiency with instability was rarely the cause of failure following cruciate-retaining total knee arthroplasty in patients with rheumatoid arthritis.