BACKGROUND The differential effects of bilateral and unilateral training on upper extremity (UE) function remain unclear. OBJECTIVE To compare the effectiveness of bilateral and unilateral training on UE function and activities of daily living (ADL) after stroke. METHODS Randomized controlled trials (RCTs) were selected for inclusion by two reviewers after searching the following databases: MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL), and KoreaMed. Methodological qualities were assessed using the PEDro scale. Effect size was estimated by calculating the standardized mean difference (SMD). RESULTS Eleven RCTs of sufficient quality were included in our meta-analysis. The effect size on UE capacity was statistically significant in favor of the Constraint-Induced Movement Therapy (CIMT) groups (SMD [fixed], g: - 0.34; 95% CI: - 0.59-0.08; p = 0.01; I2 = 0%). No other SMDs were significant. CONCLUSIONS The CIMT tasks were more effective than bilateral training with regard to increased UE capacity; however, this result should be cautiously interpreted since the evaluation tools were designed for assessment of unilateral UE function, not bilateral UE function. Further, the effect of the CIMT tasks on UE capacity was not translated into ADL. Considering the disadvantages of CIMT such as fatigue, bilateral training may be more appropriate for improving ADL.