The accurate assessment of adrenal function is necessary in many children with suspicion of pituitary insufficiency. The objective of this study was to evaluate the adrenal response during the glucagon stimulation test (GST) and its diagnostic utility in children. A total of 290 children, aged 10.1 +/- 5.0 years, were evaluated for adrenal function using the corticotrophin releasing hormone (CRH) test, the GST, and/or the insulin tolerance test (ITT). Glucagon stimulation provoked a substantial rise in cortisol and adrenocorticotropin (ACTH) that was independent of gender, age, or underlying growth hormone deficiency. There were no differences in peak cortisol levels in the GST compared to the CRH test in pair-wise intra-individual analyses in children with both tests performed within one year (615.4 +/- 30.5 vs 602.8 +/- 22.4 nmol/l, n=52). Similarly, there were no differences in the cortisol response between the ITT and CRH test. Peak cortisol levels in the CRH test correlated with the GST and the ITT. The magnitude of ACTH response, in contrast, was highest in the ITT with a 9.8-fold increase over baseline, while the increase in the GST (3.1-fold) and CRH test (1.6-fold) were more subtle. Since there is controversy concerning reliable cut-off values for adrenal function tests in children, we analyzed cut off levels in 186 children, including 26 children with adrenal insufficiency, using the CRH test. A peak cortisol level of 450 nmol/l provided the best balance of sensitivity (88.5%) and specificity (86.8%), while higher cut-off levels did not increase sensitivity but lost in specificity. In summary, the GST constitutes an1 equally sensitive test for the assessment of adrenal function in children that is not confounded by anthropometric parameters and is generally not accompanied by major side effects. It allows the simultaneous assessment of corticotroph and somatotroph function and may thus constitute a valuable alternative to the ITT.