This study attempted to characterize pharmacologically the involvement of 5-HT(2A) receptors in 5-HT-induced contractile responses in human umbilical vein (HUV) rings employing functional and radioligand binding assays. In HUV rings, prazosin 1 micro M did not affect contractile responses elicited by 5-HT, ruling out the involvement of alpha(1)-adrenoceptors in contractile responses to 5-HT. 5-HT-induced contractions were competitively blocked by ketanserin, a 5-HT(2A)-selective antagonist. The apparent pA(2) value was 9.8 and the Schild slope significantly less than unity, suggesting that 5-HT-induced responses are mediated by a heterogeneous receptor population. Alpha-methyl-5-HT, a selective 5-HT(2) receptor agonist, induced contractions that were antagonized in a competitive manner by ketanserin. The slope regression was not significantly different from unity and the pA(2) value was 8.8. The selective 5-HT(2A) ligand spiperone produced a parallel rightward shift on 5-HT CRCs in HUV rings. The calculated pA(2) was 9.0, which is in accord for an interaction with the 5-HT(2A) receptor subtype. Alpha-methyl-5-HT CRCs were competitively blocked by spiperone treatment. The Schild analysis yielded a pA(2) of 9.1 with a slope not significantly different from unity. The 5-HT(2C/2A) antagonist mesulergine 10 nM did not affect 5-HT CRCs, suggesting that 5-HT(2C) receptors are not involved in 5-HT-elicited contractions. Higher concentrations of mesulergine showed a parallel rightward shift on 5-HT responses. The calculated pA(2) was 7.44, which suggests an interaction with the 5-HT(2A) receptor subtype. In addition, mesulergine competitively blocked alpha-methyl-5-HT CRCs. The Schild slope was not significantly different from unity and the p A(2) value was 7.98. The binding of [(3)H]ketanserin to HUV membranes was saturable and of high affinity. Ketanserin displayed a monophasic curve which was best fit with a single component of binding. Nonlinear least squares analysis of the binding curves revealed a high affinity K(d) of 0.30 nM and a B(max) of 134 fmol/mg protein. These findings provide strong pharmacological evidence of the involvement of 5-HT(2A) receptors in 5-HT-induced vasoconstriction in HUV. In addition, the contribution of another receptor population cannot be excluded. The results also suggest that this receptor population is neither an alpha(1)-adrenoceptor nor a 5-HT(2C) receptor subtype.