We assessed the need for continued glucocorticoid replacement therapy in postsurgicalpituitary tumor patients using a dexamethasone-ACTH test.The patients received 1 mg dexamethasone, orally, at 2300 h and 250 ng synthetic ACTH (Cosyntropin), iv, at 0800 h the next morning. The mean ±SD integrated cortisol response for a 2-h period of the 31 pituitary tumor patients[1264 ±924 μg-min/dl (34.87 ± 25.49 μmol-min/liter)]was significantly less (P <0.005) thanthat of 25 normal subjects [3331 ±544 μg-min/dl (91.90 ±17.04 Mmolmin/liter)]. Replacement glucocorticoids were abruptly discontinued in 11 patients with responses above 1450 jig- min/dl (40.01 janol- min/liter). No clinical or laboratory evidence of adrenal insufficiency occurred as long as15 months after discontinuation. Metyrapone tests, however, in the 11 glucocorticoid-withdrawn patients revealed a reduced mean ±SD serum 11-deoxycortisol level compared with that of 10 normalsubjects [8.9 ±4.7 vs. 15.6 ±5.0 Mg/dl (0.26 ±0.13 vs. 0.45 ±0.16 x00B5;mol/liter); P < 0.005]. Ourresults indicate thatthe dexamethasone-ACTH test is useful inidentifying patients in whom replacement glucocorticoid therapy can be safely withdrawn under nonstressed conditions. The test can besimplified to one plasma cortisol level determined 30 min after ACTH administration.