Following an endplate potential there is a period of elevated spontaneous quantal release, known as delayed release. We have estimated the relationship between evoked quantal release and delayed release in two ways. First, in a solution in which the Na+ was replaced with methylamine, the number of delayed release evoked by electronic depolarization of the nerve terminal changed little over a 30-fold range of output. Second, we varied [Ca2+]out stepwise in Na solution with added Mg2+ and measured delayed release and EPP amplitude at single junctions. Over much of the range, elevating the EPP amplitude many-fold increased delayed release only slightly, if at all. Our interpretation is that the effects of residual Ca2+ are not proportional to the amount of Ca2+ entering to trigger quantal release.