The `kiss-and-run` model of exocytosis and endocytosis predicts that synapt
ic vesicles can undergo fast and efficient recycling, after fusion with the
plasmalemma,without intermixing of membranes. Evidence is mounting from se
veral new experimental approaches that kiss-and-run occurs at synapses. Dis
tinct vesicle pools, which initially were identified in morphological terms
, are now being characterized in biochemical and functional terms. In addit
ion, at least two functional recycling pathways, operating on different tim
e scales (from milliseconds to tens of seconds, have been shown to coexist
in the same synaptic system, and the two pathways appear to be differential
ly regulated. Taken together, these data suggest that kiss-and-run operates
in parallel with the classical, coated-vesicle recycling. Here, we review
recent evidence for kiss-and-run recycling and discuss whether it is a dist
inct process, dependent on the molecular organization of the fusing vesicle
. We propose that vesicles undergo a process of `competence maturation' Acc
ording to this view, the specific molecular make-up of the vesicles,their l
ocation and their interactions with nerve terminal proteins might determine
not only the differential availability of the vesicles for fusion and neur
otransmitter release but also the recycling path that they will follow.