Background and Purpose-Previous studies have shown that repeated sensory inputs could enhance brain plasticity and cortical motor output. The purpose of this study was to investigate whether combining electrically induced sensory inputs through transcutaneous electrical nerve stimulation (TENS) with task-related training (TRT) in a home-based program would augment voluntary motor output in chronic stroke survivors better than either treatment alone or no treatment. Methods-Eighty-eight patients with stroke were assigned randomly to receive a home-based program of (1) TENS, (2) TENSϩTRT, (3) placebo TENSϩTRT, or (4) no treatment (control) 5 days a week for 4 weeks. Outcome measurements included Composite Spasticity Scale, peak torques generated during maximum isometric voluntary contraction of ankle dorsiflexors and plantarflexors, and gait velocity recorded at baseline, after 2 and 4 weeks of treatment, and 4 weeks after treatment ended. Results-When compared with TENS, the combined TENSϩTRT group showed significantly greater improvement in ankle dorsiflexion torque at follow-up and in ankle plantarflexion torque at week 2 and follow-up (PϽ0.01). When compared with placeboϩTRT, the TENSϩTRT group produced earlier and greater reduction of plantarflexor spasticity and improvement in ankle dorsiflexion torque at week 2 (PϽ0.01). When compared with all 3 groups, the TENSϩTRT group showed significantly greater improvement in gait velocity (PϽ0.01). Conclusions-In patients with chronic stroke, 20 sessions of a combined TENSϩTRT home-based program decreased plantarflexor spasticity, improved dorsiflexor and plantarflexor strength, and increased gait velocity significantly more than TENS alone, placeboϩTRT, or no treatment. Such improvements can even be maintained 4 weeks after treatment ended.