Attempts to correlate 5-hydroxytryptamine (5-HT) binding sites and the receptor mediating 5-HT-induced contractions in the canine basilar artery (CBA) have led to discrepant claims, particularly regarding the potencies of antagonists (which appear to act noncompetitively) in the CBA. In the classical theoretical model for noncompetitive antagonism, the antagonist does not inhibit agonist binding. Thus attempts to correlate noncompetitive antagonist pD2' values with binding measurements are inconsistent with this model and are of questionable significance. Only agonist dissociation constants (KA values) were measured in the CBA (by the irreversible antagonist method) and correlated to binding data. Logs of these KA values did not correlate with log Ki values for the inhibition of binding of the 5-HT2 ligand [3H] ketanserin (r = 0.2253, P greater than .05) or log IC50 values for inhibition of [3H]5-HT binding (r = 0.5732, P greater than 0.05), which labels both 5-HT1A and 5-HT1B sites. A correlation was observed between the CBA values and log Ki values for the 5-HT1A binding site subtype (r = 0.9456, P less than .01). The agonist ED50 and KA values were in most cases essentially identical, which suggests that in this tissue there are no spare receptors for 5-HT. Also, identical values were obtained when apparent noncompetitive antagonists were used in a manner analogous to the irreversible antagonist method to determine KA. These findings suggest the hypothesis that the apparent noncompetitive antagonism may be produced by slow dissociation of the antagonist-receptor complex, producing a pseudoirreversible action which is unmasked due to the lack of spare receptors.