That severe diabetes mellitus is a disease of insulin deficiency has never been questioned. In contrast, for the milder forms of the disease, widely varying (and often conflicting) views have been offered in order to explain the pathophysiology of the metabolic abnormality (for a review of this topic, see 1). However, after many swings in opinion, during recent years most workers seem to acknowledge that even the mildest forms of diabetes are accompanied by deficient insulin production in response to the needs of the organism. This, together with the fact that techniques for hormone measurement and organ isolation have been available for the endocrine pancreas for a number of years, has stimulated a huge number of studies on the mechanism of insulin secretion. Indeed, any young scientist with the intention of working in this field will probably get confused when first confronted with the literature, and unless he is totally unconscious or extremely self-confident, will seek some other topic of inquiry, where simpler questions may be given simple answers. For this reason, I shall attempt in this review to offer a simplified version of how the secretion of insulin may be governed by glucose (and glucose alone, for the sake of further simplicity), and which steps in this process may be modified in diabetes mellitus. To seek for the least number of explanations that may account for the largest amount of observations (accepting the evident risk of oversimplifications) has been my leading motive.