Background: We evaluated the results of the May modification of the Bristow-Latarjet procedure ("coracoid in standing position") in 319 shoulders with respect to (1) coracoid healing and position and (2) surgical treatment of the joint capsule. Methods: From 1980 until 2004, all shoulders with a Bristow-Latarjet repair were registered at our hospital. This study consists of 3 different cohorts with respect to follow-up. Series 1, 118 shoulders operated on during 1980 through 1985, had 15 years' radiographic and clinical follow-up. Series 2, 167 shoulders that had surgery during 1986 through 1999, underwent retrospective follow-up by a questionnaire and scores-Western Ontario Shoulder Instability Index; Disabilities of the Arm, Shoulder and Hand; and Subjective Shoulder Value-after 10 to 23 years. Series 3, 34 shoulders treated during 2000 through 2004, with an added modified Bankart repair ("capsulopexy") in 33 shoulders, were prospectively followed up for 5 to 8 years with the same questionnaire and scores as series 2. Results: Of 319 shoulders, 16 (5%) had 1 or more redislocations and 3 of these (1%) had revision surgery because of remaining instability. One or more subluxations were reported in 41 shoulders (13%). The worst scores were found in 16 shoulders with 2 or more subluxations (P < .001). Radiographs showed bony healing in 246 of 297 shoulders (83%), fibrous union in 34 (13%), migration by 0.5 cm or more in 14 (5%), and no visualization in 3 (1%). Five of six shoulders that had the transplant positioned 1 cm or more medial to the glenoid rim had redislocations (83%, P = .001). Shoulders with migrated transplants did not differ from those with bony or fibrous healing with respect to redislocations and subluxations. When just a horizontal capsular shift was added to the transfer, the recurrence rate (redislocations or subluxations) decreased, with 2 of 53 (4%) compared with 37 of 208 (18%) with just anatomic closure of the capsule (P = .005), and the Western Ontario Shoulder Instability Index score improved (92 vs 85.6, P = .048). In total, for 307 of 319 shoulders (96%), patients were satisfied or very satisfied at final follow-up. Conclusion: The open Bristow-Latarjet procedure yields good and consistent results, with bony fusion of the coracoid in 83%. A position of the coracoid 1 cm or more medial to the rim meant significantly more recurrences. The rate of recurrences decreased and subjective results improved when a horizontal capsular shift was added to the coracoid transfer. (c) 2012 Journal of Shoulder and Elbow Surgery Board of Trustees.