Objective: To compare the short- and long-term effects of a structural-oriented (convential) with an activity-oriented physiotherapeutic treatment in patients with frozen shoulder. Design: Double-blinded, randomized, experimental study. Setting: Outpatient clinic. Subjects: We included patients diagnosed with a limited range of motion and pain in the shoulder region, who had received a prescription for physiotherapy treatment, without additional symptoms of dizziness, a case history of headaches, pain and/or limited range of motion in the cervical spine and/or temporomandibular joint. Interventions: The study group received treatment during the performance of activities. The comparison group was treated with manual therapy and proprioceptive neuromuscular facilitation (conventional therapy). Both groups received 10 days of therapy, 30 minutes each day. Main measures: Range of motion, muscle function tests, McGill pain questionnaire and modified Upper Extremity Motor Activity Log were measured at baseline, after two weeks of intervention and after a three-month follow-up period without therapy. Results: A total of 66 patients were randomized into two groups: The activity-oriented group (n = 33, mean = 44 years, SD = 16 years) including 20 male (61%) and the structural-oriented group (n = 33, mean = 47 years, SD = 17 years) including 21 male (64%). The activity-oriented group revealed significantly greater improvements in the performance of daily life activities and functional and structural tests compared with the group treated with conventional therapy after 10 days of therapy and at the three-month follow-up (p < 0.05). Conclusions: Therapy based on performing activities seems to be more effective for pain reduction and the ability to perform daily life activities than conventional treatment methods.