In studies designed to determine the mechanism by which Ca ++ and calmodulin stimulate the fusion of parotid secretion granules with plasma membrane vesicles, the hypothesis tested was that Ca ++ and calmodulin act by stimulating protein phosphorylation. It was earlier found that Ca ++ and calmodulin, but neither alone, stimulated the phosphorylation of four secretion granule proteins with molecular masses of 64, 58, 55 and 31 kDa, and decreased the degree of phosphorylation of a 36-kDa protein. Further studies have shown that in the presence of an optimal concentration of calmodulin (2.4 μM), half-maximal activation of phosphorylation of the four proteins occurred at approx. 8 μM Ca ++, and at a maximally effective Ca ++ concentration (10 −4M), half-maximal stimulation occurred at calmodulin concentrations between 0.13 and 1.1 μM for the different proteins. The studies now described also demonstrate that the need for calmodulin for stimulating the phosphorylation, but not the dephosphorylation, is specific; two other Ca ++-binding proteins, parvalbumin and troponin, could not replace calmodulin in stimulating phosphorylation of the four secretion granule proteins, but either one could substitute for calmodulin in stimulating dephosphorylation of the 36-kDa protein. Additionally, the phosphorylated proteins appear to be located on the granule surface. When secretion granules were subjected to mild treatment with a concentration of trypsin that did not lyse the granules, the 31-, 36−, 55−, 58− and 64-kDa proteins were no longer observed. In the presence of optimal concentrations of Ca ++ and calmodulin, a dose-dependent inhibition of the phosphorylation of the various proteins by two calmodulin antagonists, trifluoperazine and calmidazolium, was observed; 50% inhibition of phosphorylation of the different proteins was obtained at approx. 20–40 μM trifluoperazine and at about 2.5–3.0 μM calmidazolium. Inhibition of the dephosphorylation of the 36-kDa protein required greater concentrations of trifluoperazine and calmidazolium; 128 μM and 50 μM, respectively. These results are consistent with the hypothesis that the phosphorylation of one or more of the 31−, 55−, 58− and 64-kDa Proteins, but not the dephosphorylation of the 36-kDa protein, may be involved in the action of Ca ++ and calmodulin in secretion granule-plasma membrane fusion.