Frozen shoulder is said to be a self-limiting entity but full recovery often takes more than 2 years. For that, most patients are unwilling to tolerate painful restriction while awaiting resolution. We prospectively investigated 30 patients (16 women, 14 men) for the outcome of arthroscopic capsular release in idiopathic frozen shoulder. Results were determined by the assessment of subjective and objective parameters to estimate both shoulder function and general health status. Symptoms persisted without improvement for a minimum of 6 months of conservative treatment. Preoperative average American shoulder and elbow surgeons score (ASES) was 35, visual analog scale (VAS) to measure pain was 7, and simple shoulder test (SST) was 4. Mean scores of the physical component of SF-36 were considerably reduced. Mean forward elevation was 85 degrees , average abduction was 70 degrees , mean internal rotation was 15 degrees , and mean external rotation was 10 degrees . Patients were followed-up at 6 weeks, 3, 6, 12 months and by a mean of 36 months. Range of motion for all planes improved (P < 0.05). Median VAS reduced to 2, average ASES increased to 91, and SST enhanced to a mean of 10 (P < 0.05). We stated improvement of the physical components in the SF-36 questionnaire in particular bodily pain and the role-physical score. There were no significant differences between the measurements in the early postoperative phase compared to the mid-term follow-up (P > 0.05). Our results demonstrate that arthroscopic release of refractory idiopathic frozen shoulder combined with a gentle manipulation provides reliable expectations for improvement in both clinical and general health status for most patients. We recommend the use of a limb-specific and a general-health-status questionnaire to conclude the benefit of the surgical intervention and contribute the optimization of a therapy concept more effectively.