5-Hydroxytryptamine (5-HT) was applied by pressure ejection to coeliac ganglion cells of the guinea-pig maintained in vitro and responses measured intracellularly. 2. Cells responded in one of three ways to 5-HT: by (a) a fast, transient depolarization (43%), (b) a fast transient followed by a slow depolarization (biphasic response, 30%) or (c) a slow sustained depolarization (25%). 3. Fast depolarizations (response (a) above] were graded according to the duration of the ejection pulse. Maximal responses had a mean amplitude of 12 +/- 0.8 mV, a duration of 6.4 +/- 1.0 s, a latency of 0.4 +/- 0.1 s, were associated with a fall in membrane input resistance, increased in amplitude by hyperpolarization and probably mediated by an increased conductance to Na and K. The estimated reversal potential was -22.8 +/- 2.4 mV (n = 14). The maximal fast response seen in biphasically-responding cells (b) appeared similar to fast response (a). 4. Fast depolarizations (a) showed marked tachyphylaxis and were abolished by superfusion of the ganglion with 5-HT (100 microM). They were reduced in amplitude by tubocurarine (10-100 microM, pIC50 4.4), MDL 72222 (1-5 microM, pIC50 5.8), quipazine (1 microM reduced responses by 65 +/- 15%, n = 3), ICS 205-930 (1 microM reduced responses by 64 +/- 14%, n = 7) and metoclopramide (10 microM reduced responses by about 45%), but were unafected by methysergide (up to 1 microM) or hexamethonium (up to 1 mM). 5. Slow depolarizations (c) varied in amplitude with the duration of the ejection pulse. Maximal responses had a mean amplitude of 6.4 +/- 0.7 mV, a duration of 62 +/- 6 s, a latency of 3.5 +/- 0.8 s and were reduced in amplitude by methysergide (0.1-1 microM, pIC50 6.5) but not by MDL 72222 (1 microM). The maximal slow component in biphasically-responding cells (b) was similar in amplitude and duration to slow response (c), was partially blocked by methysergide (1-5 microM) in 4 of 6 cells and was enhanced by tubocurarine (50 microM) which reduced the fast component. 6. Slow depolarizations (b,c) were associated with either a small reduction or no change in membrane input resistance depending on the cell studied. Hyperpolarization had variable effects on slow depolarization amplitude. 7. It was concluded that the fast, phasic depolarization is mediated by an ionic mechanism and by receptors both of which are distinct from those involved in the slow depolarization. The receptor mediating the fast depolarization is a 5-HT3 receptor while that mediating the slow depolarization has yet to be identified.