1. The three-microelectrode voltage clamp technique and pharmacological agents were used to examine the properties and functions of potassium currents in squid giant presynaptic terminals. 2. Outward currents consisted of two components: a slow component which activated over hundreds of milliseconds and was blocked by extracellular application of tetraethylammonium (TEA) ions and a more rapidly activating component which was relatively insensitive to extracellular TEA. 3. The more rapid component was studied in isolation by treating presynaptic terminals with extracellular TEA, as well as tetrodotoxin (to block sodium channel currents) and manganese (to block calcium channel currents). The magnitude of this current component was 1-2 mA cm-2 at 0 mV. Rates of activation and deactivation were voltage dependent and little evidence of inactivation was seen for depolarizations less than several seconds in duration. 4. The reversal potential of the current was -70 to -80 mV in normal saline and became more positive with elevated extracellular potassium concentrations, suggesting that potassium is the primary permeant ion. Accumulation of extracellular potassium appeared to be marked during depolarizations that produced significant activation of the current. 5. Extracellular application of 3,4-diaminopyridine (DAP) blocked the current with an apparent dissociation constant of 7 microM at 0 mV. Intracellular applications of DAP and TEA also were effective in reducing this current. These treatments, but not extracellular TEA application, broadened presynaptic action potentials and increased the magnitude and time-to-peak of postsynaptic currents elicited by the broadened presynaptic action potentials. Postsynaptic currents were a sensitive and linear function of action potential duration; a 30% increase in action potential duration increased postsynaptic current amplitude by 190%. 6. Estimation of the magnitude and time course of the presynaptic calcium current, based on previous measurements of calcium channel gating, indicated that action potential broadening produces a large increase in calcium current magnitude. These calculations predict that a 30% increase in presynaptic action potential duration will increase the peak amplitude of the calcium current by approximately 170% and the total amount of calcium entry by approximately 230%. This implies a linear relationship between transmitter release and calcium entry during an action potential and can be explained by assuming that calcium co-operatively triggers release within intracellular domains that do not overlap.(ABSTRACT TRUNCATED AT 400 WORDS)