Studies on synaptic proteins involved in neurotransmitter release often aim at distinguishing between their roles in vesicle priming (the docking of synaptic vesicles to the plasma membrane and the assembly of a release machinery) as opposed to the process of vesicle fusion. This has traditionally been done by estimating two parameters, the size of the pool of fusion-competent vesicles (the readily releasable pool, RRP) and the probability that such vesicles are released by an action potential, with the aim of determining how these parameters are affected by molecular perturbations. Here, it is argued that the assumption of a homogeneous RRP may be too simplistic and may blur the distinction between vesicle priming and fusion. Rather, considering priming as a dynamic and reversible multistep process allows alternative interpretations of mutagenesis-induced changes in synaptic transmission and suggests mechanisms for variability in synaptic strength and short-term plasticity among synapses, as well as for interactions between short- and long-term plasticity. In many cases, assigned roles of proteins or causes for observed phenotypes are shifted from fusion- to priming-related when considering multistep priming. Activity-dependent enhancement of priming is an essential element in this alternative view and its variation among synapse types can explain why some synapses show depression and others show facilitation at low to intermediate stimulation frequencies. Multistep priming also suggests a mechanism for frequency invariance of steady-state release, which can be observed in some synapses involved in sensory processing.