Ankle arthroscopy has recently allowed the elaboration of less invasive techniques for the treatment of anterior impingement. Its indications, advantages, and drawbacks in this application are discussed. Between 1987 and 1994, 133 patients were treated for ankle impingement. Among them, 58 patients, 37 men and 21 women (mean age, 28.5 years), who had failed a trial of conservative treatment were treated by means of tibiotalar arthroscopy. Twenty-seven were athletes engaged in sports with abnormal stressing of the ankle. According to McDermott's radiological classification, there were 15 stage I cases, 23 stage II, 13 stage III, and 7 stage IV. Preoperative evaluation with a modified version of McGuire's scoring system gave 50 cases rated as "poor" (< 60 points) and 8 cases rated as "fair" (60-67 points). Treatment consisted of removal of adhesions, cartilage shaving, and removal of the bone impingement with powered instruments, curettes, or small osteotomes. Follow-up was from 8 to 62 months (mean, 21.5 months). The postoperative McGuire ratings were 37 good, 13 fair, and 8 poor. There were no major complications. Recurrence of impingement was observed in four cases of stage III and IV. The conclusion is drawn that ankle arthroscopy is a sound method for the treatment of anterior impingement. Even in cases with severe joint cartilage impairment, it plays a therapeutic role as a means of postponing a possible arthrodesis.