Facilitation at the crayfish neuromuscular inhibitor synapse was investigated with the use of a presynaptic voltage control method in which 5-ms presynaptic pulses were used to activate and monitor facilitation. A single 5-ms pulse was able to activate facilitation with a decay time constant similar to that of the F2 component of facilitation activated by action potentials. The quality of the control of presynaptic potential during F2 facilitation was evaluated by measuring the amplitude of presynaptic pulses and by analyzing the shape of the depolarization-release coupling plot during facilitation. Both approaches suggested that neither the amplitude of presynaptic depolarizations nor the space clamp of the presynaptic axon was changed during F2 facilitation. The activation of facilitation was examined by changing the amplitude of conditioning pulses systematically and using a test pulse of a constant amplitude to monitor facilitation. We found that a significant amount of facilitation could be activated by conditioning pulses that were subthreshold to the activation of transmitter release. Facilitation plateaued before the inhibitory postsynaptic potentials (IPSPs) activated by conditioning pulses reached their maximum. A double logarithm plot of facilitation magnitude against the conditioning IPSP amplitude yielded a slope of 0.34, which implies that the calcium ion cooperativity of activating facilitation is about one third of the secretion process. These findings enabled us to activate near maximal facilitation, by a burst of subthreshold conditioning pulses, without any conditioning transmitter release, and, therefore, to avoid complications associated with previous transmitter release. The detection of facilitation was examined by changing test pulse amplitude systematically to evaluate the ability of the test pulse to detect a constant level of facilitation. The magnitude of normalized facilitation decreased with increasing test pulse amplitude. The magnitude of absolute facilitation (the amplitude of the facilitated minus the control IPSP) increased with increasing test pulse amplitude. A double logarithm plot between facilitated and control IPSPs gave rise to a slope of 0.77, which suggests that the calcium cooperativity of transmitter release was decreased during facilitation.