The effects of prolonged betahistine administration were studied in institutionalized geriatric patients with particularly severe and long-standing arteriosclerotic dementia. Thirty received betahistine hydrochloride (SERC) or placebo orally in fixed dosage for six months on a double-blind basis and were followed by ward behavioral and psychometric ratings. Six others received active medication and were additionally followed by intracranial rheoencephalography (IREG). Thirty patients successfully completed the two studies. No adverse effects ascribable to the drug were encountered. The results show that betahistine caused definite, strong, and highly significant cerebral and scalp arterial vasodilation and circulatory improvement and that these caused equally definite, strong, and highly significant global improvement in the patients' dementias. Bethahistine, thus, acts in humans as a potent and efficacious cerebral and peripheral microcirculatory and arterial vasodilator which can significantly improve cerebrovascular insufficiency and any associated dementia, no matter how severe either may be and despite the possible presence of large-vessel disease. These improvements sometimes were detected within two weeks or less, developed rapidly to maxima by about 90 days, and were sustained quite well there-after by continued therapy, They may, however, be drug dependent as they regressed in two patients following withdrawal. Various mechanisms whereby a microcirculatory vasodilator can cause arterial vasodilation are considered. The evidence apparently favors that prolonged microcirculatory vasodilatation can cause secondary passive arterial responses. The possibility that the IREG contains external carotid contamination is again examined. Significant discrepancies between the scalp and IREG effects of bethahistine indicate, however, that the response of any such component was negligible within experimental error and that the component, itself, must have comprised but a small percentage of the IREG. Since the circulatory responses detected by the IREG thus arose essentially completely from the effects of betahistine on the cerebral circulation alone, this appears to be the first direct demonstration that cerebral circulatory improvement can cause improvement in mental function in patients with even severe, longstanding, and apparently clinically irreversible arteriosclerotic dementia and that such improvements can be effected pharmacologically.