1. This study aimed to investigate the 5-hydroxytryptamine (5-HT) receptors mediating contraction of ring preparations isolated from human pulmonary arteries and veins. In functional studies, the responses to 5-HT, sumatriptan, ergotamine, serotonin-O-carboxymethyl-glycyl-tyrosinamide (SCMGT), alpha-methyl 5-HT (alpha-Me) and 2-methyl 5-HT (2-Me) were studied with WAY100635, GR127935, ritanserin, zacopride and SB204070 as antagonists. 2. All agonists produced concentration-dependent contractions of human pulmonary artery and vein preparations. The order of potency (-log ECS0 values) was ergotamine (6.88) > 5-HT (6.41) > or = SCMGT (6.20) = sumatriptan (6.19) > or = alpha-Me (6.04) in the artery, and ergotamine (7.84) > 5-HT (6.96) > sumatriptan (6.60) = alpha-Me (6.56) > SCMGT (6.09) in the vein. The potency of each agonist, except for SCMGT, was greater in vein than in artery preparations. Contractile responses to 5-HT were similar in intact and endothelium-denuded preparations but responses to sumatriptan were enhanced in artery rings without endothelium. 3. GR127935 (1 nM to 0.5 microM) produced an unsurmountable antagonism of the response to 5-HT, sumatriptan, ergotamine and SCMGT. Ritanserin (1 nM to 1 microM) also reduced the maximum contractile responses to 5-HT, ergotamine and alpha-Me in artery and vein preparations without affecting those to sumatriptan and SCMGT. In endothelium-denuded preparations, surmountable antagonism of sumatriptan by GR127935 (in the presence of ritanserin) and of alpha-Me by ritanserin (in the presence of GR127935) allowed for the calculation of the apparent pK(B) values of GR127935 (9.17+/-0.11 in artery and 9.11+/-0.05 in vein) and ritanserin (8.82+/-0.09 in artery and 8.98+/-0.12 in vein). 4. WAY100635 (1 nM to 1 microM), zacopride (1 nM to 1 microM), or SB204070 (1 nM) did not significantly alter the concentration-response curves for 5-HT, sumatriptan, ergotamine, SCMGT or 2-Me in human pulmonary artery or vein thus indicating that 5-HT1A, 5-HT3 and 5-HT4 receptors are presumably not involved in the contractile response to these agonists. 5. Binding studies using selective radioligands for different 5-HT receptors could not detect the presence of 5-HT1A receptor binding in human pulmonary blood vessels whereas the 5-HT(1B/1D) radioligand [3H]-5CT significantly labelled a population of specific binding sites in both vessel types. The presence of 5-HT2A receptors could also be inferred from the level of binding of [3H]-ketanserin to membranes obtained from human pulmonary vessels, although significance could not be reached for arteries. 5-HT4 specific receptor binding was scarce in veins and absent in the case of arteries. 6. These findings indicate that the human pulmonary artery and vein have a mixed functional population of 5-HT(1B/1D) and 5-HT2A receptors mediating the contractile response to 5-HT which is consistent with results of the binding studies.