Between April 1990 and April 1994, 100 patients with a preoperative diagnosis of anterior instability underwent a diagnostic arthroscopy of the shoulder. Patients with isolated SLAP lesions were excluded from the study. Patients with multidirectional instability, bony Bankart lesions, and large Hill-Sachs lesions were also excluded. Football players and dominant arm throwing athletes were also excluded. Thirty patients remained who had post-traumatic, unidirectional, anterior instability and a repairable Bankart lesion and did not have any exclusions as noted above. All of these patients underwent an initial attempt at an arthroscopic Sure-Tac stabilization (Smith & Nephew, Andover, MA). Twenty-three patients met our criteria for a secure fixation but 7 did not. These 7 underwent an immediate arthrotomy and open Bankart repair. All of the patients were available at follow-up at an average of 47 months (range, 36 to 72 months). The patients were evaluated by the Rowe shoulder rating scale. There have been two cases of recurrent subluxation and one case of recurrent dislocation in the Sure-Tac group. All three occurred over 2 years later. All three had recurrent Bankart lesions and underwent an arthrotomy and Bankart repair. There were no cases of recurrence of subluxation or dislocation in our initial open Bankart repair group. Sure-Tac arthroscopic anterior stabilization of the shoulder can initially give good results but these results appear to deteriorate over time and increased activity of the patient. We currently do not recommend a Sure-Tac repair even in a selective group of patients with an isolated Bankart lesion. This is based on our excellent results following an open Bankart repair and a 13% recurrence rate following Sure-Tac stabilization in carefully selected patients.