Objective. Assessment of the available evidence for the effectiveness of exercise therapy to improve arm function in patients who have suffered from a stroke. Methods. A systematic search of bibliographical databases and reference checking were performed to identify publications on randomized controlled trials (RCTs) which evaluated the effect of exercise therapy on arm function in stroke patients. The methodological quality was assessed systematically by two raters, based on a standardized list of methodological criteria. Study characteristics, such as the chronicity and severity of impairment of the patient population, the amount and duration of interventions, and specific methodological criteria, were related to reported effects. Results. Thirteen RCTs were identified, six of which reported positive results on an arm function test. In five of these six studies there was a contrast in amount or duration of exercise therapy between groups. Methodological scores ranged from 5 to 15 (maximum possible score: 19 points). Conclusion. Insufficient evidence made it impossible to draw definitive conclusions about the effectiveness of exercise therapy on arm function in stroke patients. The difference in results between studies with and without contrast in the amount or duration of exercise therapy between groups suggests that more exercise therapy may be beneficial.