1. The present study was undertaken to isolate and characterize pharmacologically homogeneous populations of 5-hydroxytryptamine (5-HT) receptors from a possible mixed receptor population mediating concentration of the longitudinal muscle of rat stomach fundus. Our aim was to extend the pharmacological characterization of the 5-HT2B receptor which is reported to be expressed in this preparation. 2. To minimize spontaneous activity and any influence of circular muscle on the contractile response, narrow (1-1.5 x 20 mm) segments of mucosa-denuded longitudinal muscle were used. Under these conditions, blockade of monoamine oxidase with pargyline (100 microM for 15 min) caused a leftward displacement of concentration-effect curves for both 5-methoxytryptamine (5-MeO-T) and tryptamine. Neither pargyline nor a number of uptake inhibitors affected responses to 5-HT. 3. In pargyline pretreated preparations, the order of potency of a number of tryptamine analogues was as follows: 5-MeO-T > or = alpha-Me-5-HT > or = 5-HT > 5-carboxamidotryptamine (5-CT) > tryptamine > 2-Me-5-HT. In addition several ligands known to act as agonists at either 5-HT2A or 5-HT2C receptors including 1-m-chlorophenylpiperazine (m-CPP), Ru 24969, MK 212 and SCH 23390 were also agonists in rat fundus whilst sumatriptan, renzapride and 8-hydroxy-2-(di-n-propylamino) tetralin (8-OH-DPAT) were very weak or inactive. With the exception of 2-Me-5-HT and m-CPP, most agonists produced monophasic concentration-effect curves consistent with an interaction at a single site. High concentrations of 2-Me-5-HT evoked relaxations which were blocked by phentolamine (1 MicroM) suggesting an interaction with alpha-adrenoceptors. m-CPP often evoked biphasic concentration-effect curves with a second contractile phase which was insensitive to yohimbine at concentrations higher than required for antagonism of responses to 5-HT.4. LY 53857, methiothepin, methysergide, ritanserin and ICI 170809 were potent but non-surmountable antagonists of 5-HT in rat fundus. In contrast, several ligands behaved as surmountable antagonists with the following order of potency: rauwolscine >yohimbine = mesulergine > mianserin = SB 204070 >WY 26703 > SB 200646> pirenpirone> renzapride. DAU 6285, granisetron, spiperone, ketanserin,phentolamine and GR 127935 did not affect responses to 5-HT at concentrations up to 1 pM. The agonist and concentration independent profile of antagonism supported a single site interaction for both agonists and antagonists.5. We conclude that despite small differences concerning the enantiomeric selectivity and affinity of rauwolscine and yohimbine, the close pharmacological identity of 5-HT receptors in rat stomach fundus and the recently cloned 5-HT2B receptor is maintained. SB 200646, which demonstrates some selectivity for 5-HT receptors in rat stomach fundus, should provide a useful ligand for confirmation of this view and allow discrimination of 5-HT2B function both in vitro and in vivo.