Cellular trafficking pathways end with fusion reactions at the target. These reactions have been studied extensively for many decades, but recent studies have been particularly productive in providing new solutions to old problems, especially in some of the most complex fusion reactions, like synaptic vesicle secretion in neurons. Here, we discuss new studies that begin to merge ideas on three central questions: (A) are all releasable vesicles equally likely to undergo fusion, (B) do different fusion modes contribute to synaptic transmission, and (C) which molecular events are 'upstream' and which ones 'downstream' of SNARE complex assembly.