Dislocation arthropathy after surgical treatment of recurrent anterior dislocation of the shoulder has been a subject of discussion over the years. The objective of this prospective study was to evaluate the Bristow-Latarjet repair at 2 and 15 years after surgery with respect to arthropathy and to evaluate factors responsible for this development. At 2 and 15 years after a Bristow-Latarjet repair for recurrent anterior dislocation, we prospectively analyzed the radiographs of 115 of 118 shoulders. The majority of the patients also had radiography of the nonindex shoulder (111/118). Dislocation arthropathy was found on ordinary anteroposterior views in 46 of 115 shoulders (mild in 39, moderate in 5, and severe in 2). The anteroposterior view angulated 45 degrees from above gave the best visualization of the humeral osteophyte and, together with the axial view, doubled the incidence of moderate and severe arthropathy at follow-up (mild in 40, moderate in 11, and severe in 5). Of 38 shoulders with a postoperative position of the transferred coracoid process 2 to 4 mm medial to the glenoid rim and the screw and transplant parallel to the glenoid, 3 had moderate arthropathy (8%) compared with 8 shoulders with moderate arthropathy and 5 with severe arthropathy out of 77 shoulders with a nonoptimal position of the coracoid (17%) (P = not significant). The degree of restriction of outward rotation at 2 years did not influence the degree of arthropathy after 15 years. Global assessment of the operative result was not related to arthropathy at follow-up. When the first dislocation occurred before age 23 years, the incidence of arthropathy was significantly lower than when it occurred in age 23 years or older (P = .006). When all views were included, moderate or severe dislocation arthropathy was found in 14% of the shoulders, and a further 35% had mild arthropathy. When follow-up studies of operated series of shoulder dislocations with the Bristow-Latarjet method are done, it is important to have the same radiologic views to compare the incidence of dislocation arthropathy correctly.