Preliminary assessments of the feasibility, safety, and effects on neuronal reorganization measured with transcranial magnetic stimulation (TMS) from Constraint-Induced Movement Therapy (CIMT) of the upper extremity were made in eight cases of subacute stroke. Within fourteen days of their stroke, patients were randomly assigned to two weeks of CIMT or traditional therapy. Baseline motor performance and cortical/subcortical representation for movement with TMS were assessed before treatment. Post-treatment assessments were made at the end of treatment and at three months after the stroke. The TMS mapping showed a larger motor representation in the lesioned hemisphere of the CIMT patients as compared to the controls at the three-month follow-up assessment. The enlarged motor representation in the lesioned hemisphere for hand movement correlated with improved motor function of the affected hand, suggesting a link between movement representation size as measured with TMS and functionality. These results suggest that TMS can be safely and effectively used to assess brain function in subacute stroke and further suggest that CIMT may enhance cortical/subcortical motor reorganization and accelerate motor recovery when started within the first two weeks after stroke.