Type II diabetes mellitus may affect as many as 20% of the elderly US population. In the absence of data to support the need to maintain a specific level of glucose beyond that necessary to relieve symptoms, choice of therapy is problematic. Clearly, supervised dietary therapy for the obese type II diabetic patient represents a safe and cost-effective treatment. For those patients who fail dietary therapy because they fail to lose weight or regain lost weight, or because blood glucose levels remain high despite weight loss, further therapy must be individualized. The only rational criteria for drug treatment supportable by currently available data are (1) persistent symptoms associated with hyperglycemia, (2) ketonuria in the unstressed state, and (3) certain cases of hyperlipidemia, especially with triglyceride levels greater than 1000 mg/dl. In these clinical settings, drug therapy is necessary to eliminate symptoms, prevent development of ketoacidosis, and reduce the risk of pancreatitis, respectively. Consideration of drug therapy should also be made in the case of very elevated blood glucose levels, even in the absence of symptoms, when dehydration and risk of severe hyperosmolarity exist. The issues regarding insulin versus sulfonylureas have not been examined specifically in the elderly population. Extrapolating from published studies that generally include patients older than 65 years leads to the following conclusions: Caution regarding adverse side effects of insulin (hypoglycemia, theoretic risk of hyperinsulinemia) and sulfonylureas (hypoglycemia, drug interactions, increased risk of cardiovascular death) must be balanced against the theoretic benefit of treatment in the absence of symptoms.(ABSTRACT TRUNCATED AT 250 WORDS)