Differing from the other organs, the brain gets energy only by the oxygenation of glucose and moreover has the minimum storage of glucose. The brain is therefore especially sensitive for deficiency of glucose. The hypoglycemia may occasionally cause a fatal damage for the central nervous system. Since Wilder first reported a case of insulinoma in 1924, the neuropathological study on hypoglycemia has been scarecely reported. A) The clinical and neuropahtologicl studies on the autopsy cases of hypoglycemia. The pathological findings of the brains were investigated in twenty-eight autopsy cases of hypoglycemia. There were 10 insulin induced, 9 oral drug induced, 2 insulin and oral drug combined used, 3 insulinoma and 4 other cases of hypoglycemia. The fallowing restlts were obtained : 1. In these brains, the following four (a-d) neuropathological changes were observed, a) atrophy of nerve cells, b) scattered disappearance of nerve cells, c) loosening and d) softening. 2. The cerebral localization of the loosening were observed in the following portions ; cerebral cortex (8/28 cases, 24%), ammon's horn (7/26 cases, 27%), pallidum (6/28 cases, 21%), claustrum (4/27 cases, 15%), cerebral medulla, putamen and caudatus (5/28 cases, 18%), thalamus (4/28 cases, 14%), midbrain (3/27 cases, 11%), cerebellar cortex (2/27 cases, 7%), cerebellar medulla, pons and medulla oblongata (1/28 cases, 4%). The cerebral localization of the softening were limited to ammon's horn (5/26 cases, 19%), cerebral cortex and medulla (2/28 cases, 7%), claustrum, putamen and pallidum (1/28 cases, 4%). In the cerebral cortex, the neuropathological changes were observed generally diffused, localized laminar at the bottom of the sulci and most frequently in Brodman's third layer and other portions (the second, the second-sixth and fourth-sixth layers). In twenty-one cases, the neuropathological changes [b) c) d)] in ammon's horn were localized in Sommer's sector (19/21 cases, 90%), those in end plate, siviculum and resistant band (3/21 cases, 12%) were observed scarcely. Purkinje cells had more vulnerabilities than the nerve cells in pons and medulla oblongata. 3. The blood sugar's levels of all of the twenty-sixth cases were less than 40 mg/dl and the mean value was 30 mg/dl. The cerebral softenings were more frequently seen in the lowest blood sugar's group (0-20 ml/dl, 33%), than in the higher group (21-40ml/dl, 7%). Tho cerebral loosenings were detected in the cases who had suffered from irreversible hypoglycemia more than four days until their death, and the cerebral softenings were detected in the cases who had persistnd for more than fourteen days since the first of hyphglycemia. The astro-vascular scar (laminar cortical necrosis) were observed in the two cases who died after 91 and 365 days duration of apathy. The cerebral softening was seen more frequantly in the cases with convulsions (3/26 cases, 50%), than the cases without convulsion (2/21 cases, 10%). Nevertheless, having unconsciousness for a long time, the sugar's cerebral loosening and sof tening were not found in the three cases of relatively high blood level (30-40ml/dl) and noconvulsion. There were no cerebral softening and only two cerebral loosening in the cases treated with insulin, whose irreversible hypoglycemic attacks mostly occured at the hospital. The pathological chahges of the brain of these cases were considered to mostly influenced by the duration of hypoglycemia and repeated convulsions. The selective cerebral vulnerabilities on the hypoglycemia were mostly suitable for those of anoxia. B) The experimental study on hypoglycemic effects of the rabbit's brain. Fifty-four rabbits were divided to four groups; group I (oral administration of tolbutamide), group II (oral abministration of tolbutamide on alloxan diabetic rabbits), group III (alloxan initial hypoglycemia), group IV (control).