The incidence of type 2 diabetes mellitus (DM) in children and adolescents has substantially increased over the past decade. The present study was conducted to evaluate the beta-cell response to intravenous glucagon (a non-glucose secretagogue) in children with type 2 DM. Twenty pediatric patients with type 2 DM were compared to 15 control subjects matched for body mass index and sexual maturation. The patients' ages ranged between 10 and 19 years. The duration of DM ranged from 1 to 5 years. Nine patients were on insulin treatment and 11 were on diet alone (3 patients) or metformin (8 patients). The criteria for type 2 DM were absent islet cell (IA-2) and glutamic acid decarboxylase (GAD65) antibodies and a fasting serum C-peptide level of > or = 0.23 nmol/l. Plasma glucose and serum C-peptide levels were determined in the fasting state and six minutes after an intravenous injection of 1 mg of glucagon. The fasting and stimulated plasma glucose levels and the fasting serum C-peptide levels (1.02 +/- 0.43 vs 0.79 +/- 0.26 nmol/l, p < 0.05) were higher in the patients with DM compared to weight-matched control subjects. While the absolute C-peptide responses to glucagon were not different between the two groups, the stimulated C-peptide to glucose ratios were significantly lower in the patients with DM compared to controls (0.039 +/- 0.026 vs 0.062 +/- 0.033, p < 0.05). Patients with DM treated with diet or oral therapy had significantly greater basal and stimulated C-peptide concentrations, incremental C-peptide, and C-peptide to glucose ratios than patients on insulin treatment. Both the fasting and the stimulated C-peptide levels were inversely correlated with the duration of DM (r = -0.53, p < 0.05). HbA1c at one year follow-up was inversely correlated with glucagon-stimulated C-peptide levels at the time of the study (r = -0.658, p < 0.01) and positively correlated with the duration of diabetes (r = 0.671, p = 0.002). The apparently normal serum C-peptide levels measured after glucagon challenge in these children with type 2 DM reflect their higher glucose levels. The lower stimulated C-peptide to glucose ratios in these children with type 2 DM compared to normal controls demonstrate their diminished beta-cell response to intravenous glucagon, a non-glucose secretagogue. Among the patients with DM, a higher glucagon-stimulated serum C-peptide response was associated with diet/metformin treatment, a shorter duration of DM and predicted improved glycemic control up to one year later. Thus, the fasting and glucagon-stimulated serum C-peptide levels provide an estimate of the potential insulin secretory capacity of the beta-cell and may predict glycemic control in pediatric type 2 DM.