Background: Many treatment strategies have been described for the management of frozen shoulder. However, to date no randomized controlled trial has directly compared the efficacy of exercise and oral corticosteroids. The aim of this study was to determine whether pain, range of motion (ROM), and function differ between patients with frozen shoulder randomized to treatment with exercise or oral corticosteroids. Methods: This prospective, randomized, assessor-blinded trial was performed to compare the effectiveness of exercise and oral corticosteroids in patients with frozen shoulder. Overall, 33 patients with frozen shoulder were randomly assigned to receive either exercise (n = 17) or oral corticosteroids (n = 16). The exercise was applied 2 times per week for 6 weeks for 12 visits; patients in the oral corticosteroids group received prednisolone for 4 weeks. Participants were assessed at baseline, after 6 weeks and at the 12-week follow-up. The primary outcome was the Disabilities of the Arm, Shoulder and Hand and visual analog scale. Secondary outcomes were American Shoulder and Elbow Surgeons Standardized Shoulder Assessment, ROM, and Hospital Anxiety and Depression Scale. Repeated-measures analysis of covariance with baseline scores as the covariates was used to determine between-group differences. An intention-to-treat analysis was performed using the multiple imputation method to impute values for all missing data. Results: Planned pairwise comparisons demonstrated significant improvements in pain relief and functional outcomes in both groups at the 6- and 12-week follow-ups compared to baseline. There were no significant time-by-group interactions between-group differences noted for the Disabilities of the Arm, Shoulder and Hand (F = 0.470, P = .93), visual analog scale (F = 0.006, P = .94), flexion ROM (F = 2.78, P = .1), internal rotation ROM (F = 3.440, P = .07) and American Shoulder and Elbow Surgeons Standardized Shoulder Assessment (F = 0.470, P = .49). The overall group-by-time interaction for the 2 × 3 mixed-model analysis of covariance using baseline scores as a covariate was significant for the abduction range (F = 4.460, P = .04) and external rotation ROM (F = 12.100, P = .002) in favor of the exercise group. Conclusions: The study demonstrated that while both groups achieved significant improvements at the 6-, and 12-week follow-ups, the exercise group was superior in terms of abduction and external rotation ROM. Additionally, even though both groups were improved, the effect sizes were larger in the exercise group. Considering the systemic side effects of oral corticosteroids, even at low doses, a well-planned exercise program that considers pain may be a good option for frozen shoulder.