1. Ionic currents associated with the invasion of an action potential into the motor nerve ending of the lizard, Anolis carolinensis, were measured with a focal extracellular electrode at several locations along the nerve ending. 2. These experimentally observed currents could be matched with computer simulations of action potential propagation into the nerve ending. They revealed that while Na+ channels are the major ionic current pathway in the heminode, K+ channels provide the major pathway in the terminal branches and boutons. 3. Calcium current in the presynaptic ending was unmasked by the application of tetraethylammonium (TEA). This current was blocked by: (a) cadmium, (b) omega-conotoxin GVIA and (c) nifedipine, but was unaffected by nickel at concentrations less than or equal to 100 microM. Nifedipine's action became more definitive when the duration of the action potential was greatly extended by pre-treatment with TEA. The effect of Bay K 8644 was inconsistent. 4. Transmitter release, as measured by postsynaptic current, had a pharmacological response profile similar to that of the Ca2+ current, with the exception that transmitter release was increased reliably and reversibly by Bay K 8644. 5. This pharmacological response profile is identical to that of the L type Ca2+ channel identified by Fox, Nowycky & Tsien (1987 alpha) in chick dorsal root ganglion neurones. We saw no evidence for more than a single type of Ca2+ channel in lizard motor nerve endings. 6. A calcium-activated K+ current IK(Ca) was revealed by application of 3,4-diaminopyridine (DAP), a delayed-rectifier K+ channel blocker. This K(Ca) current was blocked by TEA, charybdotoxin and by substitution of cobalt for extracellular calcium.