Purpose: The clinical outcome of the revision surgery for the failed Bankart repair is not well known. The purpose of this study was to prospectively analyze a series of patients with recurrent instability after primary Bankart repair that were revised arthroscopically using a suture-anchor technique. Type of Study: Prospective, nonrandomized outcome study. Methods: Twenty-three patients with failed Bankart repair were treated with arthroscopic surgery and were followed up for a mean of 36 months (range, 24-52 months; SD, 9 months). There were 21 men and 2 women with the mean age of 24 years (range, 17-34 years; SD, 4.4 years). Eight patients had previously received an open Bankart repair (5 transosseous-suture technique, 3 suture-anchor technique), and 15 had received an arthroscopic repair (10 transglenoid multiple-suture technique, 5 suture-anchor technique). The revision surgery included repair of the anterior labrum using suture anchors and nonabsorbable sutures, capsular plication, and proximal shift of the inferior capsule with or without closure of the rotator interval. The characteristics of the patients, possible modes of failure, surgical findings, shoulder scores (University of California at Los Angeles [UCLA] scale, Simple Shoulder Test, and Rowe score), and clinical outcome were evaluated. Results: Recurrent instability developed at a mean of 21 months (range, 11-39 months; SD, 8 months) after the initial stabilization. A possible cause of failure in 5 patients was a nonanatomic repair, with the labral tissue fixed proximal or medial to the glenoid margin. At the follow-up, 15 patients had excellent results, 4 good, 3 fair, and 1 poor, according to the UCLA scale. The mean Rowe score improved from 87.3 (range, 30-100; SD, 12.6) to 91.2 (range, 40-100; SD, 14.2) (P =.023). The Simple Shoulder Test responses improved from 8 yes responses preoperatively to 11 postoperatively. Eighteen patients achieved an activity return of more than 90% of preinjury level. Five patients had recurrence after revision surgery (1 frank dislocation, 2 subluxation, 2 positive anterior apprehension sign). Engagement in contact sports was correlated with the recurrence (r =.683, P =.003). With the number available, no significant difference could be detected in the outcome between the types of the primary surgery. Arthroscopic revision Bankart repair did not result in any additional loss of external rotation compared with the contralateral side (P =.723). The most predictable factor for the functional return was preoperative range of external rotation (r =.793, P =.001). Conclusions: Arthroscopic revision Bankart surgery using suture anchors can provide a satisfactory outcome, including a low recurrence rate and reliable functional return, in carefully selected patients.Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 18, No 5 (May-June), 2002: pp 469–482