Purpose The purpose of this study was to clinically validate the Hill–Sachs interval to glenoid track width ratio (H/G ratio) compared with the instability severity index (ISI) score for predicting an increased risk of recurrent instability after arthroscopic Bankart repair. Methods A retrospective evaluation was performed using data from patients with anteroinferior shoulder instability who underwent arthroscopic Bankart repair with a follow-up period of at least 24 months. A receiver operating characteristic (ROC) curve was used to determine the optimal cut-off values for the H/G ratio and the ISI score to predict an increased risk of recurrent instability. The area under the ROC curve (AUC) of the two methods and the sensitivity and specificity of their optimal cut-off values were compared. Results A total of 222 patients were included, among whom 31 (14.0%) experienced recurrent instability during the follow-up period. The optimal cut-off values for predicting an increased risk of recurrent instability were an H/G ratio of ≥ 0.7 and ISI score of ≥ 4. There were no significant differences between the AUC of the two methods (H/G ratio AUC = 0.821, standard error = 0.035 and ISI score AUC = 0.792, standard error = 0.04; n.s.) nor between the sensitivity and specificity of the optimal cut-off values (n.s. and n.s., respectively). Conclusions The H/G ratio is comparable to the ISI score for predicting an increased risk of recurrent instability after arthroscopic Bankart repair. Surgeons are recommended to consider other strategies to treat anterior shoulder instability if H/G ratio is ≥ 0.7. Level of evidence III.