Abstract: Objective To investigate whether the constraint-induced movement (CIM)therapy is more effective than traditional upper-extremity therapies after ischemic stroke. Methods Forty patients with ischemic stroke involving persistent hemiparesis, evidence of preserved cognitive function, and presence of a protective motor response, persons were enrolled in a pilot randomized, controlled trial that compared CIM with traditional therapies for 10 days, 2 hours a day. Patients in the CIM group wore cotton pad gloves for at least 6 hours a day during the 10-day treatment period, encouraging the patients to use their hemiplegic arms to complete daily tasks. Patients in the traditional treatment group received standard rehabilitation treatment. Differences of ARAT, Barthel Index and FIM between the groups were compared by using student's t tests, constituent ratio by using Chi square tests. Results No subject withdrew because of pain or frustration. The CIM treatment group had significantly higher scores on total ARAT and pinch subscale scores (all P<0.05). Differences in the mean ARAT grip, grasp, and gross movement subscale scores did not reach statistical significance. Conclusions A clinical trial of CIM therapy during acute rehabilitation is feasible. CIM was associated with less arm impairment at the end of treatment. Long-term studies are needed to determine whether CIM early after stroke is superior to traditional therapies.