Background : Constraint-induced movement therapy (CI therapy) is a rehabilitation treatment that effectively improves upper extremity function in patients with chronic hemiparesis after stroke. In Japan, no previous studies have shown the effects of CI therapy in a large size sample, and the Simple Test for Evaluating Hand Function (STEF), a standard assessment tool for upper extremity function, is rarely used to evaluate the effects of CI therapy. Little is known regarding the factors capable of predicting the outcome of CI therapy. The present study aimed to examine potential predictors of outcomes after CI therapy and the effects of CI therapy using STEF. Methods : This study included patients with hemiparesis in the chronic stage of stroke (≥180 days from onset). We compared upper extremity function before and after CI therapy intervention, which involved 5 hours per day for 10 consecutive weekdays of training. We assessed upper extremity function using the STEF, Wolf Motor Function Test-functional ability scale (WMFT-FAS) before and after intervention, and investigated potential predictors (age, gender, time since stroke, type of stroke, side of stroke, hand dominance, spasticity). Results : There were 107 subjects. Our comparison revealed that both STEF and WMFT-FAS scores improved significantly following intervention, from means of 31.3 to 42.7 points and 51.8 to 57.0 points, respectively. No significant predictors that influenced the functional outcome were identified. Conclusion : CI therapy is effective for improving upper extremity function. It is important not to exclude people from CI therapy based on any supposed predictors.