PURPOSE: To evaluate by means of MR arthrography the causal relationship between anterior capsularattach-ment type and anterior instability of the glenohumeral joint. MATERIALS AND METHODS: Sixty-five patients (M:F=64:1; mean age, 23 years), who had undergone shoulder MR arthrography were retrospectively reviewed. There wasno history of traumatic dislocation, and medical records relating to anterior instability were available. Physicalexamination revealed anterior instability in 25 patients(Group A), while in 40 (Group B), this was eithersuspicious or absent. With the aid of general anesthe-sia, 32 patients who had undergone surgery were physicallyexamined. Anterior instability was evident in 16 patients(Group C), and suspicious or absent in the remaining 16(Group D). All images were obtained after in-traarticular injection of Gd-DTPA. Fat-suppressed T1 axial imagingwas used to evaluate anterior capsular at-tachment type at mid and distal three-fourths levels of the glenoidfossa. The causal relationship between ante-rior capsular attachment types and anterior instability was analyzedusing the x2 test RESULTS: In group A, type I was most common at distal three-fourths level, and types I and IIwere equal and more common than type III at mid level. In group B, type I anterior capsular attachment type wasmost com-mon at both mid and distal three-fourths level. In these groups, no statistical significance was notedbetween anterior capsular attachment type and anterior instability (p value 0.611: 0.567). In group C, types I andII were equal and more common than type III at both mid and distal three-fourths level, while in group D, type IIwas most common at mid level, and type I was most common at distal three-fourths level. In these groups, nostatistical significance was noted between anterior capsular attachment type and anterior instability (p val-ue,0.772). CONCLUSION: There was no statistically significant difference between anterior capsular attachment typeand anterior instability of the shoulder joint.