Recent experimental evidence suggesting that presynaptic depolarization can evoke transmitter release without calcium influx has been re-examined. The presynaptic terminal of the squid giant synapse can be depolarized by variable amounts while recording presynaptic calcium current under voltage clamp and postsynaptic responses. Small depolarizations open few calcium channels with large single channel currents. Large depolarizations approaching the calcium equilibrium potential open many channels with small single channel currents. When responses to small and large depolarizations eliciting similar total macroscopic calcium currents are compared, the large pulses evoke more transmitter release. This apparent voltage-dependence of transmitter release may be explained by the greater overlap of calcium concentration domains surrounding single open calcium channels when many closely apposed channels open at large depolarizations. This channel domain overlap leads to higher calcium concentrations at transmitter release sites and more release for large depolarizations than for small depolarizations which open few widely dispersed channels. At neuromuscular junctions, a subthreshold depolarizing pulse to motor nerve terminals may release over a thousand times as much transmitter if it follows a brief train of presynaptic action potentials than if it occurs in isolation. This huge synaptic facilitation has been taken as indicative of a direct effect of voltage which is manifest only when prior activity raises presynaptic resting calcium levels. This large facilitation is actually due to a post-tetanic supernormal excitability in motor nerve terminals, causing the previously subthreshold test pulse to become suprathreshold and elicit a presynaptic action potential. When motor nerve terminals are depolarized by two pulses, as the first pulse increases above a certain level it evokes more transmitter release but less facilitation of the response to the second pulse.(ABSTRACT TRUNCATED AT 250 WORDS)