Purpose: To investigate the effect of adding cognitive strategy training to task-specific training (TST), called Cognitive Oriented Strategy Training Augmented Rehabilitation (COSTAR), compared with TST on activity and participation for chronic stroke survivors in an outpatient occupational therapy setting Materials and methods: We conducted an exploratory, single-blind, randomized controlled trial. Participants were randomized to TST or COSTAR protocol. Our primary outcomes measured activity and participation after stroke: the Stroke Impact Scale (SIS), Canadian Occupational Performance Measure (COPM), and Performance Quality Rating Scale (PQRS). Results: Forty-four participants were randomized. The COSTAR group had an attrition rate of 50% and an average of 9.8 of 12 sessions were completed; the TST group had an attrition rate of 25% and an average of 10.7 sessions were completed. Generally both groups improved on the majority of primary and secondary outcomes. There is little evidence to support a beneficial effect of COSTAR over TST for improvement of primary measures of activity performance or secondary measures. Conclusion: Negligible findings may be attributed to an inadvertent treatment group equivalency. Further, the research design did not allow for adequate measurement of the effect of each intervention on participants' ability to generalize learned skills. Implications for rehabilitation Stroke rehabilitation is largely based upon the principles of task-specific training, which is associated with improvements in upper extremity motor performance; however, TST requires a heavy dosage and lacks generalization to untrained activities. Cognitive strategy use has been associated with improved generalization of treatment to untrained activities and novel contexts however, it is often not used in TST protocols. The results of this preliminary study found no clear advantage between task-specific training and strategy-adapted task-specific training on trained and untrained activities when both interventions targeted activity performance. Task-specific training, if focused at the activity performance level rather than the impairment reduction level, may have a stronger effect on improving in individual's ability to participate in everyday life activities even without the use of cognitive-strategies. Incorporating cognitive strategy-use into TST would likely produce the greatest effect on generalization and transfer of the treatment effects to other activities and contexts rather than solely on activity performance of trained activities.