The insulin tolerance test (ITT) is considered the test of choice for the diagnosis of GH deficiency (GHD). However, in patients with contraindications to ITT, alternative provocative tests must be used with appropriate cut-offs. The glucagon stimulation test has proved to be a safe, low-cost and effective means of stimulating GH secretion, and therefore can be considered as a suitable alternative to the ITT. We have studied the GH response to the glucagon test in 33 patients with known pituitary disease, 12 males and 21 females, aged between 21 and 60 yr (41.18 +/- 9.47 yr); 5 had isolated GHD and 28 had panhypopituitarism. We also evaluated a control group of 25 individuals, matched for age and sex (8 males and 17 females), aged between 20 and 60 yr (39.28 +/- 12.10 yr). They were selected via the ITT if their peak GH response was > 5.0 ng/ml. GH peak after glucagon was significantly lower in the group of patients compared to the control group (0.49 +/- 0.85 vs 8.69 +/- 5.85 ng/ml, p = 0.0001). Receiver-operating characteristic (ROC) plot analyses of the control and GHD group showed an area under the curve of 0.982 for GH peak response to glucagon. The response value of 3.0 ng/ml showed the best pair of sensitivity (97%)/specificity (88%), and was chosen as the cut-off defining GHD. After evaluation of positive predictive values (PPV) and negative predictive values (NPV) through simulation of different prevalences of the disease, we concluded that the cut-off point of 3.00 ng/ml maximizes both PPV and NPV (100%). In conclusion, we have shown that the glucagon stimulation test has a good performance and great diagnostic accuracy for the diagnosis of GHD.