The current controversy concerning the role of glucagon in the pathogenesis of diabetes is reviewed. The traditional "unihormonal abnormality concept," namely, that all of the metabolic derangements of diabetes are the direct consequence of deficient insulin secretion or activity, and the newer so-called bihormonal abnormality hypothesis, proposing that the fullblown diabetic syndrome requires, in addition to the insulin abnormality, a relative glucagon excess, are scrutinized. The relationship of insulin deficiency to the A-cell malfunction of diabetes, the conflicting evidence concerning the essential role of glucagon in mediating the marked overproduction of glucose and ketones in severe insulin deficiency and the contribution of glucagon to the endogenous hyperglycemia of diabetics without insulin deficiency are examined. Finally, the possibility that therapeutic suppression of diabetic hyperglucagonemia may make possible better control of hyperglycemia than is presently attainable by conventional therapeutic methods is considered. It is concluded that (1) although insulin lowers glucagon levels, restoration to normal of the A-cell dysfunction of diabetes requires that plasma insulin levels vary appropriately with glycemic change; (2) that glucagon mediates the severe endogenous hyperglycemia and hyperketonemia observed in the absence of insulin; (3) that in diabetics in whom insulin is present but relatively fixed an increase in glucagon causes hyperglycemia and glycosuria; and (4) that glucagon suppression could be a potentially useful adjunct to conventional antihyperglycemic treatment of diabetics.