PurposeRecurrent shoulder instability, a common musculoskeletal disorder, often involves glenoid bone loss and Hill–Sachs lesions. However, the optimal imaging modality for accurately and reliably quantifying bipolar bone loss remains uncertain. This systematic review aims to evaluate the accuracy and reliability of various imaging modalities in assessing bipolar bone loss in anterior shoulder instability.MethodsMajor electronic databases were searched for English‐language studies reporting the measurement of glenoid track width and/or determination of on/off track HSL through imaging. Studies reporting statistical measures such as area under the curve, sensitivity, specificity, positive predictive value, NPV, intraobserver reliability and interobserver reliability were included. Data extraction and risk of bias assessment were performed independently by two reviewers.ResultsThe systematic review included 19 studies comprising 1567 shoulders. Overall, studies could be divided into those looking at the accuracy or reliability of determining glenoid track width, on‐ or off‐track Hill–Sachs lesions and near‐track lesions. Three‐dimensional images of computed tomography (3D‐CT) was the most reliable and accurate imaging modality to measure the glenoid track width. On the opposite, two‐dimensional magnetic resonance imaging (2D‐MRI) did not provide enough evidence of accuracy and reliability in the determination of On/Off track lesions and near‐track lesions.Conclusion3D‐CT demonstrated excellent reliability for measuring glenoid track width. However, the reliability of 2D‐MRI for determining on/off track Hill–Sachs lesions is still controversial.Level of EvidenceLevel III.