Ever since 1952 when Delay, Deniker and Heal first used chlorpromazine for various psychiatric patients, its clinical effect has come to be recognized. And as is wellknown it is now one of important drugs in the treatment of patients with mental illness. Believing that the clarification of the mechanism acting on the brain will still further the pathological study in endogenous psychoses, the author performed a series of experiments with dog, in which he studied clinical symptoms and side-effecects at the time of chlorpromazine administration and also carefully analyzed the results of histopathological findings on the brain for the purpose of elucidation of the mechanism of chlorpromazine acting on the brain. Namely, dogs were divided into two groups: the A-group given a large dose of chlorpromazine for a short period of time; and the B-group given a small does of chlorpromazine for a long time. 1. For the A-group, grown-up dogs and young dogs were selected to the total of nine dogs, and in order to give shockwise 43-133 mg/kg chlorpromazine was injected into the artery, vein or muscle. As the result the clinical stage can be divided into five stages: 1. somnolent stage; 2. lethargic stage; 3. paralytic stage; 4. dyspnea stage; and 5. agonal stage. All of them died within several days. It was revealed that various symptoms of motor disturbances were most apt to appear in the lethargid stage. Especially the young dog No.8 showed a marked torsion dystonia-like symptoms at this atage and these symptoms persisted thereafter. All of them were sacrificed by decapitation six hours after the injection, and removing and fixing the brains, histological specimens were prepared. 2. For the B-group two adult dogs were selected and 10 mg/kg chlorpromazine was injected every day intramuscularly. Although clinical symptoms could not be divided into different stages, there was a period when they became somnolent. There were decapitated 31 days after the start of experiment and tissue specimens were prepared in the same way as mentioned above. By comparing there clinical symptoms in dogs with those observed in hyman cases, the author studied the mechanism of chlorpromazine action in the brain. In the histopathologisal investigations specimens were stained with hematoxylin-eosin, azo-carmine, Nissl stain, myelin sheath stain, and fat stain (Sudan III). In the A-group changes changes appearing diffusely in the entire brain, the so-called acute changes, and those of nerve cells caused bue to the changes in blood vessels were recognized. Sites especially marked for such changes were in the corpus striatum, thalamus, putamen, globus pallidum, nucleus niger, nucleus ruber, corpus mammillaris, nucleus amygdae and a portion of cerebral cortex. In comparison of these changes with pathoanatomy of torsion dystonia in the literatures, the changes mentioned above seemed to substantiate torsion dxstonia-like symptoms. In the B-group their characteristic changes were chronic atrophy of nerve cells and demyelinating plaques in cortical medullary radiation. Moreover, these changes were found to appear selectively at a definite portion of the brain; and it is believed that these findings offer the clue for the acting mechanism of chlorpromazine in the hrain. Finally for the purpose of explaining summarily the acting mechanism of chlorpoomazine the author discussed first the sites of chlorpromazine action, its distribution in the brain, and then the acting mechanism of chlorpromazine reported in available literatures, and also made a comparison between the results obtained by other investigators and those in the present experiment as well as the correlation with clinical symptoms.