We had previously shown that an influx of extracellular Ca2+ (Ca2+e), though it occurs, is not strictly required for aminoethyldextran (AED)-triggered exocytotic membrane fusion in Paramecium. We now analyze, by quenched-flow/freeze-fracture, to what extent Ca2+e contributes to exocytotic and exocytosis-coupled endocytotic membrane fusion, as well as to detachment of "ghosts"-a process difficult to analyze by any other method or in any other system. Maximal exocytotic membrane fusion (analyzed within 80 msec) occurs readily in the presence of [Ca2+]e > or = 5 x 10(-6) M, while normally a [Ca2+]e = 0.5 mM is in the medium. A new finding is that exocytosis and endocytosis is significantly stimulated by increasing [Ca2+]e even beyond levels usually available to cells. Quenching of [Ca2+]e by EGTA application to levels of resting [Ca2+]i or slightly below does reduce (by approximately 50%) but not block AED-triggered exocytosis (again tested with 80 msec AED application). This effect can be overridden either by increasing stimulation time or by readdition of an excess of Ca2+e. Our data are compatible with the assumption that normally exocytotic membrane fusion will include a step of rapid Ca(2+)-mobilization from subplasmalemmal pools ("alveolar sacs") and, as a superimposed step, a Ca(2+)-influx, since exocytotic membrane fusion can occur at [Ca2+]e even slightly below resting [Ca2+]i. The other important conclusion is that increasing [Ca2+]e facilitates exocytotic and endocytotic membrane fusion, i.e., membrane resealing. In addition, we show for the first time that increasing [Ca2+]e also drives detachment of "ghosts"-a novel aspect not analyzed so far in any other system. According to our pilot calculations, a flush of Ca2+, orders of magnitude larger than stationary values assumed to drive membrane dynamics, from internal and external sources, drives the different steps of the exo-endocytosis cycle.