When blood glucose is the major source of the carbons of liver glycogen, it would be expected that, on being measured with C14 label, the specific activities of the glucose and glycogen would be similar. This proved to be the case when large quantities of C14 glucose were administered orally to fasted normal and adrenalectomized rats. However, when trace quantities were given intravenously and the liver glycogen and blood glucose specific activities were determined during a 4-hr period after cortisol administration, less than 10% of the carbons of the glycogen formed could be attributed to the circulating glucose. Mixtures of glucose-6-H3, glycerol-C14 and alanine-Cu were given orally to fasted rats. Absorption was established by measurement of residual reducing substance and radioactivity in the intestine. It is estimated that about 50% of the carbons of the glycogen formed were derived from glycerol carbons, 20% from glucose, 5–10% from alanine and the remainder from endogenous sources. Cortisol administration increased the quantity of glycogen formed, but did not change its relative sources. Thus, cortisol administration does not result in a deposition of glycogen carbons primarily derived from glucose; rather, the source of the carbons seems dependent upon the substrates available. The action of cortisol would appear to be to increase the pathway of glycogen formation without preferentially stimulating any of the several pathways by which glycogen carbons are derived. It is suggested that a small increase above normal in the quantity of glucose-6-phosphate converted to glycogen could account for glycogen deposition from acute cortisol administration. Such a mechanism need not require an increase in the rate of glucose-6-phosphate production by the liver.