1. 5-Hydroxytryptamine (5-HT) stimulated a biphasic increase in short-circuit current (SCC) in guinea-pig isolated ileal mucosa. The initial 'spike' response to 5-HT was inhibited by tetrodotoxin (0.3 microM). We have investigated the 5-HT receptor mechanism(s) controlling the second 'maintained' component of the response which remained after treatment with tetrodotoxin. 2. 5-HT stimulated concentration-related increases in SCC with an EC50 value of 5.4 microM. Isobutyl-methylxanthine (IBMX, 10 microM) produced a six fold leftward shift of this concentration-response curve, suggesting the involvement of a cyclic nucleotide(s) in these responses. 3. In the presence of IBMX, 5-HT stimulated reproducible increases in SCC with an EC50 value of 0.9 microM. The rank order of potency of indole agonists in these tests was 5-HT greater than or equal to 5-methoxytryptamine greater than 5-carboxamidotryptamine = alpha-methyl-5-HT much greater than 2-methyl-5-HT. 4. The substituted benzamides were partial agonists. Metoclopramide and cisapride produced approximately 20% of the 5-HT maximum, and renzapride and R,S-zacopride produced approximately 50% of the 5-HT maximum. Metoclopramide and cisapride inhibited the SCC responses to 5-HT with apparent pKB values of 4.8 and 7.0 respectively. 5. The SCC responses to 5-HT were not inhibited by antagonists selective for 5-HT1 (methysergide, methiothepin), 5-HT2 (ketanserin) or 5-HT3 (ondansetron, ICS205-930) receptors. 6. The SCC responses to 5-methoxytryptamine, 5-carboxamidotryptamine, alpha-methyl-5-HT and R,S-zacopride, but not 5-HT, were selectively inhibited by high concentrations of ICS205-930 with apparent pKB values of approximately 6.7. A possible interpretation of these results is that the 'maintained' SCC response to 5-HT is mediated by a heterogeneous population of 5-HT receptors. One of these receptors exhibits the characteristics of the putative 5-HT4 receptor.