We describe a model of synaptic transmitter release and presynaptic facilitation that is based on activation of release sites by single Ca2+ C microdomains. Facilitation is due to Ca2+ that remains bound to release sites between impulses. This model is inherently stochastic, but deterministic equations can be derived for the mean release. The number of equations required to describe the mean release is small, so it is practical to use the model with models of neuronal electrical activity to investigate the effects of different input spike patterns on presynaptic facilitation. We use it in conjunction with a model of dopamine-secreting neurons of the basal ganglia to demonstrate that transmitter release is greater when the neuron bursts than when it spikes continuously, due to the greater facilitation generated by the bursting impulse pattern. Finally, a minimal form of the model is described that is coupled to simple models of postsynaptic receptors and passive membrane to compute the postsynaptic voltage response to a train of presynaptic stimuli. This form of the model is appropriate for neural network simulations.