In summary, we have presented evidence which relates to the action pathway of hormonal control of glycogen metabolism. In the case of insulin, there are changes demonstrable in the cyclic AMP-dependent protein kinase and also in the phosphoprotein phosphatase, under conditions where no direct relationship to either cyclic AMP or cyclic GMP levels are measurable. Therefore, a new unknown intermediate or second messenger system is again proposed. An insulin-generated labile compound(s) which inhibits the protein kinase has been discovered. This may function as an intermediate. Finally, the fact that the glycogen synthase system clearly differs from phosphorylase in its regulation by covalent phosphorylation is discussed. Synthase is now accepted as a multiply phosphorylated subunit, in contrast to phosphorylase which is singly phosphorylated. The inherent theoretical advantages of multiple phosphorylation over single phosphorylation are considered. The advantages of a multistate over a two-state model of enzyme interconversion are mentioned. The importance of the multiple phosphorylations interacting in a nonlinear manner with the control by cellular metabolites is in the explanation of how a small change in covalent phosphorylation signalled by a hormone can be translated in the cell milieu into a much larger change in rate.