The receptors responsible for contraction to serotonin (5-HT) in the canine basilar artery have not been definitively established to date. Several selective 5-HT2 receptor antagonists (spiperone, ketanserin and LY53857) did not inhibit markedly 5-HT-induced contractions in the canine basilar artery in doses higher than required for substantial inhibition of 5-HT2 receptor-mediated responses. These data suggest that the receptors mediating 5-HT-induced contractions in the basilar artery are not 5-HT2 receptors. Using a series of 5-HT antagonists with relatively high affinity at 5-HT1 sites, over a 1000-fold difference occurred in their ability to block 5-HT receptors in the canine basilar artery, in spite of the similar and high affinity of the antagonists at 5-HT1 binding sites. These data support the contention that 5-HT receptors in the canine basilar artery are not 5-HT1 receptors as defined by ligand binding studies in brain cortical membranes. Similarity of the contractile effects of 5-HT in the rat stomach fundus and in the basilar artery coupled to the previous observations that receptors mediating 5-HT-induced contractions in the fundus were not 5-HT1, 5-HT1A, 5-HT1B or 5-HT2 led us to consider the possibility that 5-HT receptors in the canine basilar artery may resemble those in the rat stomach fundus. The affinity of several 5-HT antagonists determined in the canine basilar artery correlated extremely well (correlation coefficient = 0.96) with the affinities obtained for the same antagonists in the rat stomach fundus.(ABSTRACT TRUNCATED AT 250 WORDS)