At many synapses, the amount of transmitter released by action potentials increases progressively during a train of spikes. This enhancement of evoked transmitter release grows during tetanic stimulation with several time constants, each bearing a different name (facilitation: tens to hundreds of milliseconds; augmentation: several seconds; potentiation: several minutes), and the enhancement of release to test spikes after a tetanus decays with similar time constants. All these processes depend on presynaptic Ca2+ influx during the conditioning tetanus. It has often been proposed that these forms of synaptic plasticity are due to residual Ca2+ present in nerve terminals following conditioning activity. We tested this idea directly by using photolabile Ca2+ chelators to reduce residual Ca2+ following conditioning stimulation or to generate an artificial elevation in Ca2+ concentration, and observed the effects on synaptic transmission at crayfish neuromuscular junctions. We found that facilitation, augmentation and potentiation are caused by the continuing action of residual Ca2+. Augmentation and potentiation seem to arise from Ca2+ acting at a separate site from facilitation, and these sites are different from the molecular target triggering neurosecretion.