The purpose of our study was to compare the ability of intraluminally and extraluminally administered nimodipine to inhibit serotonin-induced cerebral vascular responses in vitro and in situ. No difference was noted in the ability of nimodipine, whether administered intraluminally or extraluminally, to reduce the contractile response of extraluminally administered serotonin in a closed, pressurized, in vitro bovine middle cerebral artery preparation; histologic studies indicated that the tight endothelial junctions normally found in cerebral arteries remained intact in this preparation. In cats, pretreatment with nimodipine did not significantly reduce the ability of intracisternally injected serotonin to decrease cerebral blood flow; however, nimodipine did reduce the changes in cerebral artery diameter normally noted angiographically after serotonin injection. Although minor differences were noted between the intraluminal and extraluminal routes of administration of nimodipine in situ, in general the effects were comparable.