Setting: Tertiary children’s hospital. Patients: 5 children with hemiplegia and learned nonuse of the affected upper extremity. Case Descriptions: Patients included a 3-year-old boy with left hemiplegia from right schizencephaly, 4-year-old girl and 6-year-old boy with right hemiplegic cerebral palsy, 5-year-old girl with right hemiplegia following viral encephalitis at age 3 weeks, and 8-year-old boy with left hemiplegia 15 months after traumatic brain injury. All children had wrist extension of at least 20° and digit extension of at least 10° in the affected extremity. The constraint-induced movement therapy (CIMT) intervention in all children included provision of either a solid or bivalved cast for the unaffected upper extremity for a duration of 2 or 4 weeks, to force use of the affected extremity. For 4 children, 3 daily sessions of forced practice were provided as a home exercise program. Activities used were individually selected following baseline evaluation by the research occupational therapist and modified during the child’s usual weekly occupational therapy visit. The length of each daily practice session was 15 minutes for 1 patient, 30 minutes for 1 patient, and 1 hour for 2 patients. Measures at baseline and postintervention included standardized subtests of the Bruininks-Oseretsky Test of Motor Proficiency and Peabody Developmental Motor Scales, the Jebsen Hand Function Test, and videotaping. Assessment/Results: In all cases, individual motor function trajectories showed improvements in the chronically affected upper extremity greater than those expected by development over the period of intervention. Discussion/Conclusion: This exploratory evaluation of modifications in the standard adult CIMT protocol for children with hemiplegia suggests that varied protocols of CIMT can be effective. Further prospective evaluation of variations in protocols for CIMT in pediatric hemiplegia with more subjects, controls, and longer-term follow-up is indicated to identify the most therapeutically and cost-effective intervention.