PURPOSE: We evaluated the outcomes of lateral retinacular release (LRR) after a long-term follow-up period of 5 to 12 years. TYPE OF STUDY: Long-term retrospective clinical follow-up study. PATIENTS AND METHODS: Between 1986 and 1994, 120 LRRs were performed in the Orthopaedic Department of the Catholic University of Rome. A total of 100 patients were evaluated. We divided the patients into 2 groups: group I contained 50 patients with patellar pain and no signs of instability; the remaining 50 patients, with clear signs of patellar instability, made up group II. Standard weight-bearing radiographs, axial views of the knee at 45 degrees , and dynamic computed tomography scans were performed in all patients preoperatively and at follow-up evaluation. Chondral damage was classified at the time of lateral release according to the criteria of Outerbridge and Dunlop. We used the Lysholm II score, which was modified for patellofemoral pathology and a clinical grading system of Busch and de Haven, to evaluate clinical outcomes at follow-up evaluation. RESULTS: In group I (pain), 70% reported satisfactory outcomes at follow-up evaluation compared with 50% in group II (P < .05) (instability). Compared with a previously published analysis of 3-year outcomes in this same patient population, there was very little change in group I patients, whereas group II showed a significant decrease in good outcomes over time. The worst results were obtained in cases with serious cartilage damage and exposure of the subchondral bone at the time of lateral release. CONCLUSIONS: LRR is a procedure offering a good percentage of success in the management of a stable patella with excessive lateral pressure and elective location of pain on the lateral retinaculum. In patellar instability the results are less favorable in long-term follow-up evaluation. The presence of high-grade joint surface injury is a poor prognostic indicator for lateral release.