Quantitative and semiquantitative methods have been used to evaluate the clinical improvement of 62 patients with completed stroke who were admitted to a rehabilitation hospital. Improvement in motility and leg strength on the paretic side was minimal and was not influenced by facilitation exercise techniques. Observed changes in strength and motility occurred to about the same degree on both the nonparetic and paretic sides. Patients who had a short interval between onset of stroke and admission to the rehabilitation program improved significantly more on the paretic side than those with a longer interval. Practically no improvement in motility and leg strength was found two months following a stroke. Hemiplegia and hemiparesis were defined quantitatively on the basis of motility test scores. Patients with hemiparesis showed greater improvement in motility and self-care status in about half the time of hospital stay when compared with patients with hemiplegia. In spite of relatively static neurological deficits, all patients showed evidence of functional improvement as assessed by a self-care rating scale. The poorest functional outcome was seen in patients who had hemisensory losses in addition to hemiplegia. The observations indicate that early, functionally oriented stroke rehabilitation programs offer the best chance of aiding patients. Behavioral and sociological influences on final outcome are important and must be carefully evaluated to insure maximum chances of successes in rehabilitation.