Considering pancreatic reserve recognition a more rational basis for starting insulin therapy in NIDDM, during 1988 we studied fasting and post-breakfast plasma C-peptide levels in 31 patients (21 w, 10 m, mean age 48.2 +/- 17.7 yr), referred to our department for insulin therapy evaluation because of primary or secondary failure to other measures. Major features were obesity and chronically uncontrolled illness. Our patients were categorized as follows: group A, considered non responders which included four patients; group B, taken as responders consisting in seven; and a remaining of 20 hyperresponsive patients which formed group C; these with patients of group B, embodied an 87.1% of patients. Among nonresponders there was no any case of total B cell loss of function, and plasma C-peptide activity surpassed through those limits considered for ketoacidosis. We believe that these patients should be eligible candidates for insulin therapy. We failed to found out correlation of plasma C-peptide activity with either age or duration of illness. Our observation supports that fasting plasma C-peptide evaluation would suffice for pancreatic reserve evaluation. We conclude that our patients mainly presented an insulin resistant state associated with obesity thus enhancing the commonplace call for reinforcing nonpharmacologic treatment modalities such as caloric restriction, weight loss and exertion to achieve a better control in NIDDM patients.