Neuromuscular stimulation may facilitate motor recovery after stroke or brain injury, reduce shoulder pain associated with hemiplegia, and reduce cerebral spasticity. However, the discomfort of surface neuromuscular stimulation significantly limits the clinical implementation of this modality for persons with hemiplegia. The study contained herein tests the hypothesis that stroke and brain injury survivors with chronic hemiplegia (>6 mo) and intact sensation tolerate percutaneous intramuscular stimulation better than surface stimulation. Four stroke and two traumatic brain injury survivors participated in the study contained within this article. Each subject received three pairs of percutaneous and surface stimulations of the paretic finger extensors. The order of the type of stimulation within each pair was randomly assigned. The stimulation parameters for each type of stimulation were normalized to produce the same torque at the metacarpophalangeal joint. Subjects rated their perceived level of discomfort using a 10-cm visual analog scale and the McGill Pain Questionnaire. A blinded evaluator administered the pain measures. Percutaneous stimulation was associated with significantly lower discomfort as reflected by the visual analog scale (0.74v3.3; 95% confidence interval of difference, −3.84, −1.28). The McGill Pain Questionnaire produced similar results with percutaneous stimulation associated with a significantly fewer number of words chosen to describe the discomfort (0.87v3.30; 95% confidence interval of difference, −3.50, −1.30) and significantly lower Pain Rating Index (1.47v6.27; 95% confidence interval of difference, −7.77, −1.83). Data suggest that percutaneous intramuscular stimulation is significantly better tolerated than surface stimulation and that percutaneous stimulation may enhance patient compliance with neuromuscular stimulation treatments.