Introduction: Shoulder pain is recognized as a disabling problem. The most common causes of shoulder pain in primary care are reported to be rotator cuff disorders, acromioclavicular joint diseases, and glenohumeral joint disorders. The final diagnosis in case of chronic shoulder pain is based on a collective clinical as well as radiological evaluation which includes radiographs, ultrasonography (USG), and magnetic resonance imaging (MRI). Materials and Methods: After obtaining ethical clearance, 85 cases of chronic shoulder pain were enrolled for the study. A detailed history with clinical examination was done. Patients were subjected to X-ray anterior-posterior and axial as initial investigation. On viewing, the X-ray next modality was decided. All those cases where no obvious bony lesion was seen were further evaluated by USG and MRI. The diagnosis was confirmed by arthroscopy. Result: For partial thickness tear of supraspinatus, USG had a sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of 60%, 97.6%, 95.5%, 74.1%, and 80.3%, respectively, as compared to 88.6%, 96.0%, 93.9%, 92.3%, and 92.9%, respectively, for the same parameters on MRI. For full thickness tear of supraspinatus, USG had a sensitivity, specifi city, PPV, NPV, and accuracy of 95.2%, 90.6%, 76.9%, 98.3%, and 91.8%, respectively, as compared to 95.2%, 98.4%, 95.2%, 98.4%, and 97.6%, respectively, for the same parameters on MRI. Conclusion: As far as comparative evaluation of USG and MRI, except for full thickness tear where both the modalities had equal sensitivity, for all the other diagnoses MRI showed a higher sensitivity. However, for partial thickness tear, USG had a higher specifi city as compared to MRI; for all the other diagnoses, MRI showed a higher specifi city. In terms of overall accuracy, MRI had a higher accuracy as compared to USG, for all the diagnoses except for full thickness tears.