Our current notions of different granule pools, granule interaction with the plasma membrane, and ultimately granule and plasma membrane soluble N-ethylmaleimide-sensitive-factor attachment protein (SNARE) interactions, result largely from inferences based upon biochemical alterations of secretion kinetics. Another view of events comes from studies using total internal reflection fluorescence microscopy (TIRFM) to investigate granule behaviour immediately adjacent to the plasma membrane. The motions of secretory (chromaffin) granules in bovine chromaffin cells visualized by TIRFM are highly restricted, as if granules are caged or tethered. These small motions are regulated by ATP and Ca2+, two factors that increase priming of the secretory response. There is no evidence that granules decrease their motion immediately before secretion. To the contrary, there is a tendency for granules to increase their motions and travel within a few hundred milliseconds of fusion. Hence, the notion of a long-lived docked state as a prelude to fusion does not encompass the physical reality of molecular scale motions, multiple tethering states and significant travel immediately preceding the exocytotic event. Increased travel may increase the probability of granules interacting productively with the plasma membrane constituents, thereby, increasing the probability of fusion.