Of 300 consecutive knees (238 patients) that had undergone arthroplasty with the cementless Natural Knee prosthesis from 1985 to 1989, 176 knees (141 patients) were available for followup at an average of 12 ± 1 years after the operation. Knee function was improved significantly. Modified Hospital for Special Surgery knee scores improved from 59.1 ± 13.2 points preoperatively to 97.8 ± 4.7 points at last followup. At last followup, knee range of motion averaged 0° ± 2° to 120° ± 10°. Implant survival was 93.4% (including infection and simple polyethylene exchanges) and 95.1% (excluding infection and simple polyethylene exchanges) at 10 years when applying the Kaplan-Meier survival analysis, using loose components, revision, or both as failure criteria. Besides the three revisions for infection, only two femoral and one tibial component required revision. The patellar component survivorship at 10 years was 95.1%. All patellar revisions were attributed to edge wear. Subsequent operative and design changes, including patellar component medialization and countersinking, have decreased the incidence of patellar revision. The long-term results of this cementless knee system compare favorably with those of cemented systems. The Natural Knee design has provided excellent and predictable long-term clinical results in the current series of active patients.