Patients with Cushing’s disease are known to present a variable secretory response to stimulatory and inhibitory challenges. Evaluation of the secretory behaviour of pituitary adrenocorticotrophic hormone (ACTH)-secreting adenomas in vitro aids in the comprehension of its behaviour in vivo; however, given the small size of these tumours and the consequent paucity of material available to in vitro studies, a comprehensive study on the secretory behaviour of human corticotroph tumours has not yet been performed. The present study aimed to assess the spectrum of responses to the two main corticotroph modulators, corticotrophin-releasing hormone (CRH) and dexamethasone, in a large series of human ACTH-secreting pituitary tumours. Seventy-two ACTH-secreting pituitary tumours were collected during surgery and established in culture. Specimens were incubated with 10 nm CRH and/or 10 nm dexamethasone for 4 h and 24 h. Secretion in unstimulated, control wells was set at 100% and changes in ACTH concentrations by at least 20% were considered as responses. Parallel experiments in 12 rat anterior pituitary primary cultures were evaluated. A marked ACTH increase was observed during incubation with CRH in 70% of tumoural specimens at 4 h (range 124–3500% of control wells) and in 57% at 24 h (range 122–3323%). Dexamethasone reduced ACTH secretion in almost 50% of tumours (range 78–2% of control at 4 h; 76–3% at 24 h), whereas it did not affect ACTH medium levels in 30% of specimens and induced a paradoxical ACTH increase in 20% of tumours (range 130–327% of control at 4 h; 156–348% at 24 h). By comparison, CRH uniformly increased ACTH levels in rat anterior pituitary primary cultures (mean 745 ± 84% at 4 h; 347 ± 25% at 24 h), whereas dexamethasone decreased ACTH levels by 40–50% in all experiments. In conclusion, the present study of a large series of human ACTH-secreting pituitary tumours in vitro revealed a considerable variability in the responses to CRH and dexamethasone. This finding indicates the existence of multiple corticotroph tumoural phenotypes and may account for the different responses to physiological and pharmacological modulators in vivo.