The inhibitor of the crayfish opener muscle was investigated by a presynaptic voltage control method. Two microelectrodes were inserted into the inhibitor and the amplitude and duration of presynaptic depolarization were controlled by a voltage-clamp amplifier. The inhibitory postsynaptic potential (IPSP) was measured from a muscle fiber located near the presynaptic voltage electrode. Nonlinear summation of IPSP amplitudes was corrected after chloride equilibrium potential was measured. With the use of 5-ms presynaptic pulses, the depolarization-release coupling (D-R) curve constructed from IPSP peak amplitudes (IPSPcor) had a threshold of about -35 mV and reached its maximal level at -5 to -10 mV. Depolarization beyond the maximum led to a suppression of neurotransmitter release. When transmitter release during a presynaptic pulse was completely suppressed, IPSPs activated by tail current could be identified with an average synaptic delay of 2.5 ms. Transmitter secretion triggered by a calcium current activated during the 5-ms pulses (IPSPon) was also measured on the rising phase of an IPSP, at 2.5 ms after the end of the 5-ms pulses. D-R coupling plots measured from IPSPon exhibited a more pronounced suppression than that obtained from IPSPcor. The effect of presynaptic pulse duration on the level of transmitter release was analyzed. Transmitter release increased with increasing duration and was nearly saturated by 20-ms pulses depolarized to 0 mV. The following conditions were identified as necessary to obtain a consistent D-R curve with a clear suppression: 1) small animals, 3.8 cm head to tail, 2) 15 degrees C, 3) 40 mM tetraethylammonium and 1 mM 4-aminopyridine, 4) an extracellular calcium concentration of < or = 10 mM. In addition, a consistent correlation was found among the branching pattern of the inhibitor, the placement of the presynaptic electrode, and the characteristics of the D-R curves. An ideal presynaptic electrode configuration involved placing the voltage electrode in a secondary branch, approximately 100 microns from the main branch point, and placing the current electrode at the branch point. Postsynaptically, optimal recordings were obtained from muscle fibers innervated by a single branch of the inhibitor that originated from a point near the presynaptic voltage electrode. A cable-release model was constructed to evaluate the relationship between the shape of the D-R coupling curves and the space constants of the presynaptic terminals. A comparison between the model and the D-R coupling curves suggested that the space constant of an inhibitor branch on a muscle fiber is > or = 8 times longer than its actual length. Therefore the upper limit estimate of the space constant of a typical preparation is approximately 3 mm. Results reported here outline morphological and physiological conditions needed to achieve optimal control of the presynaptic branch of the crayfish inhibitor. The cable-release model quantitatively defines the extent of presynaptic voltage control.