The shoulder-hand syndrome as a clinical entity was established by Steinbrocker in 1947, and in 1958 Moskowitz made the first comment on the importance of this syndrome in stroke rehabilitation. Since that time, however, there have been only few studies on the incidence and pathogenesis of this syndrome in hemiplegia. In his paper, we performed a clinical study on the shoulder-hand syndrome in hemiplegic patients in several hospitals. The incidence of shoulder-hand syndrome in cases with post-stroke hemiplegia was 21.0per cent. Both sexes were equally affected. The influence of the age were as follows: the mean age of the patients with this syndrome was 63.7, which was considerably higher than that of patients without it (mean age 59.9). It was interesting that all the patients with this syndrome were forty years or above of age and, this syndrome was found at the highest incidence in the 6th decade. The shoulder-hand syndrome occurred at the affected upper extremity of hemiplegic patients and there was no difference of incidence between the right hemiplegia and the left. The latent period from the onset of hemiplegia to that of this syndrome ranged from the minimum of one week to the maximum of ten months, with the mean value of about 2.3 months. Our studies revealed a close relationship between the incidence of this syndrome and the Brunnstrom's recovery stages of motor function, demonstrating significantly higher incidence of this syndrome in severe hemiplegia with recovery stage from I to III for both upper and lower extremities. Another interesting finding of our studies was that this syndrome in the right hemiplegics had a tendency to be complicated with aphasia. There have been numerous theories concerning the pathogenesis of shoulder-hand syndrome, including sympathetic reflex theory and so on, and is no unaminity at the present time. We have tried to clarify some part of its pathogenesis with the mean of photoelectric volume plethysmography. Then, it was found that not all the hemiplegics, but those with shoulder-hand syndrome showed significant reduction of peripheral vascular resistance and increase of blood flow in the hemiplegic upper extremities. These findings may suggest that the increase of blood flow of hemiplegic upper extremities is either the cause of shoulder-hand syndrome, or its most important predisposing factor.