There is still uncertainty about what is the most appropriate test for assessment of the integrity of the hypothalamo-pituitary-adrenal (HPA) axis. Many advocate the insulin tolerance test (ITT), but this is unpleasant and resource intensive, and may occasionally give misleading results. The conventional [250 microg tetracosactrin, ACTH-(1-24)] short synacthen test (SST) has been used as a simple alternative to the ITT, but it has produced some falsely reassuring results with potentially serious consequences. A low dose [1 microg tetracosactrin, ACTH-(1-24)] short synacthen test (LDSST) has recently been advocated as a more reliable and safer alternative to ITT. Some studies, however, have failed to demonstrate any difference between SST and LDSST. The purpose of this study was to assess the clinical usefulness of LDSST compared to SST and ITT in patients with pituitary disease. We studied 64 patients with suspected or proven pituitary disease. All patients underwent SST and LDSST. Forty-two patients underwent ITT. There was a high correlation between the ITT and LDSST peak cortisol responses (r = 0.89; P < 0.0001), the ITT and SST 30 min cortisol levels (r = 0.83; P < 0.0001), and the LDSST peak cortisol response and the SST 30 min cortisol level (r = 0.85; P < 0.0001). In the LDSST, all but six patients achieved maximal cortisol response by 30 min. A plasma cortisol cut-off of 600 nmol/L is more helpful than 500 nmol/L for clinical decision-making using either the SST 30 min cortisol level or the LDSST peak cortisol response. The sensitivity of the LDSST was 100% (cortisol response of >600 nmol/L indicates intact HPA axis), with no falsely reassuring results. SST (pass cortisol level, >600 nmol/L) was less sensitive than LDSST, it produced 2 of 64 (3%) falsely reassuring results. Even the ITT (pass cortisol level, >500 nmol/L) failed to identify one patient with clinically evident cortisol deficiency. The results of this study indicate that both SST and LDSST, at a cortisol cut-off of 600 nmol/L, are safe for the purpose of clinical decision-making with regard to steroid replacement therapy in patients with pituitary disease. As the LDSST produced no falsely reassuring decisions, we suggest that this could replace the SST and ITT for initial evaluation of the HPA axis in patients with pituitary disease. We suggest administering 1 microg tetracosactrin, i.v., with sampling at 0, 20, and 30 min.