BACKGROUND: We prospectively compared the response to blind and ultrasound-guided glenohumeral injection of corticosteroids for treatment of shoulder stiffness. METHODS: A total of 77 patients with shoulder stiffness between April 2008 and March 2012 were recruited. Patients were randomized to receive either a blind (group 1, n=39) or ultrasound-guided (group 2, n=38) glenohumeral injection of 40 mg triamcinolone. The clinical outcomes and shoulder range of motion (ROM) before injection, at 3, 6, and 12 months after injection and at the last follow-up were assessed. The same rehabilitation program was applied in both groups during the follow-up period. RESULTS: There was no significant difference in demographic data on age, sex, ROM, and symptom duration before injection between groups (p>0.05). There were no significant differences in ROM including forward flexion, external rotation at the side, external rotation at 90o abduction, and internal rotation, visual analogue scale for pain and functional outcomes including American Shoulder and Elbow Surgeons score, Simple Shoulder test between the two groups at any time point (p>0.05). CONCLUSIONS: Based on the current data, the result of ultrasound-guided glenohumeral injection was not superior to that of blind injection in the treatment of shoulder stiffness. We suggest that ultrasound-guided glenohumeral injection could be performed according to the patient's compliance and the surgeon's preference. Once familiar with the non-imaging-guided glenohumeral injection, it is an efficient and reliable method for the experienced surgeon. Ultrasound could be performed according to the surgeon's preference.