Action potentials and afterpotentials were recorded with a microelectrode inserted into lizard motor axons within a few millimeters of their motor terminals. In the presence of 1 mM 4- aminopyridine (4-AP), the duration of the action potential recorded near motor terminals was Ca-sensitive: repolarization was more rapid when bath [Ca] was elevated, and became slower when bath [Ca] was removed or when 0.1–1 mM Mn was added. Repolarization was also slowed following addition of 3–10 nM charybdotoxin or 100–300 microM tetraethylammonium (TEA) to the bath, and following intra-axonal injection of the Ca buffer BAPTA. These results, in agreement with published extracellular recordings, indicate that the motor nerve terminal membrane contains rapidly activating, Ca-activated K channels. When these (and other) K channels were blocked by 10 mM TEA, the action potential recorded near motor terminals was followed by Ca-dependent depolarizing afterpotentials, followed in turn by a slow hyperpolarizing afterpotential (h.a.p.) that lasted several seconds. This slow h.a.p. was also Ca-sensitive: it became larger with increasing bath [Ca] and was abolished by removal of bath [Ca] and by addition of 1 mM Mn. Intra-axonal injection of BAPTA reduced the amplitude of the slow h.a.p., and prolonged injections promoted repetitive discharge. The slow h.a.p. following single action potentials was observed in 100 microM ouabain and in K-free solutions and thus is pharmacologically distinct from the hyperpolarization that follows tetanic stimulation. The slow h.a.p. was selectively inhibited by 100 nM apamin, but persisted in 100 nM charybdotoxin. This afterpotential was enhanced by 0.1–1 mM 4-AP and by the dihydropyridine Bay K 8644 (0.1–1 microM). These results suggest that the slow h.a.p. in lizard motor nerve terminals is mediated by Ca-activated K channels that can be activated near the resting potential and are pharmacologically distinct from the Ca-activated K channels that contribute to action potential repolarization. The slow h.a.p. was enhanced by 0.1–1 mM caffeine and inhibited by 100 microM procaine, raising the possibility that this afterpotential may be activated not only by Ca entering via the plasma membrane, but also by Ca released from intra-terminal stores.