BACKGROUND: Recent evidence has demonstrated the efficacy of Virtual Reality (VR) for stroke rehabilitation nonetheless its benefits and limitations in large population of patients have not yet been studied. OBJECTIVES: To evaluate the effectiveness of non-immersive VR treatment for the restoration of the upper limb motor function and its impact on the activities of daily living capacities in post-stroke patients. METHODS: A pragmatic clinical trial was conducted among post-stroke patients admitted to our rehabilitation hospital. We enrolled 376 subjects who had a motor arm subscore on the Italian version of the National Institutes of Health Stroke Scale (It-NIHSS) between 1 and 3 and without severe neuropsychological impairments interfering with recovery. Patients were allocated to two treatments groups, receiving combined VR and upper limb conventional (ULC) therapy or ULC therapy alone. The treatment programs consisted of 2 hours of daily therapy, delivered 5 days per week, for 4 weeks. The outcome measures were the Fugl-Meyer Upper Extremity (F-M UE) and Functional Independence Measure (FIM) scales. RESULTS: Both treatments significantly improved F-M UE and FIM scores, but the improvement obtained with VR rehabilitation was significantly greater than that achieved with ULC therapy alone. The estimated effect size of the minimal difference between groups in F-M UE and FIM scores was 2.5 ± 0.5 (P