Investigations in our laboratory have shown that the activity of glycogen synthase phosphatase in the liver is shared by at least two functionally distinct proteins: a G-component, which is tightly associated with glycogen particles, and a soluble S-component. Most preparations of glycogen synthase-b that are isolated from the liver of fed glucagon-treated animals require the presence of both components in order to be converted to synthase-a. The G-component is subject to control mechanisms that do not affect the S-component. Its activity is strongly inhibited by phosphorylase-a. This feature explains why glycogen synthesis and glycogenolysis do not normally occur simultaneously, except in the glycogen-depleted liver, where a futile cycle may occur. Experiments in vitro have shown that a minimal glycogen concentration is required to ensure the interaction between the G-component and phosphorylase-a. The G-component is also selectively inhibited by Ca2+, and the magnitude of this inhibition depends markedly on the glycogen concentration. The latter inhibition is probably one of the mechanisms by which cyclic adenosine monophosphate (cAMP)-independent glycogenolytic agents achieve the inactivation of glycogen synthase in the liver. Glucocorticoid hormones and insulin are required for the induction and/or maintenance of the G-component in the liver. During the development of the fetal rat, glucocorticoids induce the G-component in the liver. This is an essential event in the glucocorticoid-triggered deposition of glycogen in the fetal liver. A functional adrenal cortex is also required in the adult animal to prevent a loss of the capacity for hepatic glycogen storage during starvation. The latter capacity depends on the concentration of functional G-component in the liver. Chronic diabetes causes a similar functional loss. However, the effect of glucocorticoids is not mediated by a putative secretion of insulin.