1. Excitatory postsynaptic potentials (e.p.s.ps) were recorded from the submandibular parasympathetic ganglia of newborn rats (10-20 days old), by intracellular microelectrode recording and a suction electrode to deliver stimulus trains to the lingual nerve (15 stimuli at 0.1, 0.3, 1, 3, and 10 Hz, 22 degrees C). Only evoked responses without voltage-dependent action potentials were analyzed (observed at membrane potentials negative to -70 mV), and e.p.s.p. amplitudes were determined for the plateau responses during each train (5-15th response). 2. Cadmium, an inorganic calcium channel antagonist, reduced e.p.s.p. amplitudes in a dose-dependent manner (Kd 74 microM, P less than 0.01). Nickel (1-300 microM) did not attenuate the amplitude of evoked responses. 3. Verapamil (0.1-30 microM), a phenylamine, had no significant effects upon e.p.s.p. amplitudes at any frequency examined. Higher concentrations of verapamil (100 microM) blocked neurally evoked responses in a manner consistent with the antagonism of voltage-sensitive sodium currents. 4. Diltiazem, a benzothiazepine, reduced e.p.s.p. amplitudes in a dose-dependent manner, the depression being accentuated at high stimulation frequencies (80% block at 30 microM and 10 Hz). The pure (-)-cis enantiomer of diltiazem (10-30 microM) was without effect. 5. Amlodipine, a 1,4-dihydropyridine, did not antagonize synaptic transmission at any stimulus frequency examined (10-30 microM, 0.1-10 Hz, n = 3). 6. Amiloride, a potassium-sparing diuretic, depressed the amplitudes of evoked responses in a dose-dependent manner (one-site Kd 31 microM, P less than 0.005), although the extent of the block was alleviated with high stimulus frequencies. The effects of 30 microM amiloride were unlikely to be of post-synaptic origin as both the amplitudes of miniature e.p.s.ps, and the iontophoretic potentials induced by exogenous acetylcholine, were not attenuated by treatment with this compound. The amiloride derivative, 3',4'-dichlorobenzamil was ineffective in reducing the amplitude of e.p.s.ps (30-100 microM). 7. omega-Conotoxin GVIA, a marine neurotoxin, which depressed whole cell calcium currents recorded from cultured rat parasympathetic cardiac neurones (up to 90% block at 10 nM), was ineffective at blocking synaptic transmission in submandibular ganglia (0.1-1 microM). 8. The differential effects of these calcium channel antagonists upon synaptic transmission in rat parasympathetic ganglia, suggest that either more than one type of calcium channel may be involved in transmitter release, or that the presynaptic calcium channels possess pharmacological sensitivities different from those of channel types described in ne