Fasting plasma cyclic adenosine monophosphate (cAMP) was measured in 50 mature-onset diabetic patients and in 111 non-diabetic patients. Methods used to determine plasma cAMP are described. The addition of sepharose agar beads to the bovine adrenocortical binding protein has considerably improved the sensitivity and simplified the radioligand-receptor assay of cAMP. No statistical differences in plasma cAMP were noted in relation to sex in either group, to the presence of diabetes mellitus or to age or weight in the non-diabetic patients. Plasma cGMP levels are now being studied to determine if these may prove better indicators of insulin activity than plasma cAMP.