Purpose: The current meta-analysis reviewed evidence for effective task-oriented training focused on repetitive or circuit training after stroke. Method: Searches were conducted of randomized, controlled trials using task-oriented training. The quality of each study was assessed using the Physiotherapy Evidence Database (PEDro) scale. Eleven studies were analyzed regarding the magnitude of effect sizes (ESs) and categorized according to extremities focused upon for training, outcome measures, and study variables. This included duration and frequency of training and stroke stage. Results: The PEDro scores ranged from 4 to 8 (median57). The overall ES of the 11 studies was large. The effects for lower extremities and both acute/sub-acute and chronic stage stroke were large and significant. Significant effects were also found for gait velocity, gait endurance, balance, timed up and go test, and strength of the lower extremities. There was no significant effect of training focused on upper extremities. Training with a duration of 2 weeks and frequency of 7 days a week had the greatest effects. Conclusions: Task-oriented training interventions are useful for improving muscle strength and gait related activities in both acute/sub-acute and chronic stroke patients. Although this meta-analysis provides evidence of task-oriented training for improving functioning after stroke, further studies are necessary to investigate the effects of training on upper extremities and the overall cost-effectiveness of such training. [ABSTRACT FROM AUTHOR]