OBJECTIVES: Anecdotal reports have suggested that silent corticotroph tumours behave in an aggressive fashion; however, clear comparative data with other non-functioning adenomas (NFAs) are lacking. The aims of the study were, first, to review the natural history of those non-functioning pituitary adenomas with positive immunoreactivity for ACTH and secondly, to determine whether this subgroup behave more aggressively than ACTH immunonegative NFAs by means of comparison with existing departmental data. METHODS AND PATIENTS: Twenty-eight patients (16 men, mean age 51.3 years) who underwent transsphenoidal surgery in Oxford between 1975 and 2001 for clinically non-functioning adenomas where the subsequent immunostaining was positive for ACTH were identified from the patient database. All patients with silent corticotroph tumours who presented during this time period have been included in the analysis; three of the patients have subsequently died but none have been lost to follow-up. The mean follow-up period was 7.4 years (range 0.5-26.9 years) and the results were compared with departmental data for NFAs which were immunonegative for ACTH. None of the patients had clinical evidence of Cushing's syndrome. Tumour invasiveness was classified according to the modified Hardy criteria (Grade 1 = microadenoma (< 1 cm), Grade 2 = macroadenoma (> 1 cm) +/- suprasellar extension, Grade 3 = local invasion with bony destruction and tumour in sphenoid/cavernous sinus and Grade 4 = central nervous system (CNS) spread or extracranial spread, i.e. metastatic). Tumour recurrence was defined as an increase in tumour size compared with the first postoperative scan which was used as a baseline. Visual field defects were documented in 79% of the 28 patients at presentation compared to 69% in the non-functioning population as a whole (P = 0.3). The preoperative imaging in the silent corticotroph group (13 CT, 14 MRI and one air encephalogram) revealed 68% Grade 2 and 32% Grade 3 adenomas. RESULTS: The recurrence rate in the ACTH immunopositive tumours was 32% at a mean of 5.8 years (range 1-16 years) which was not significantly different from the 33% recurrence rate previously recorded in the ACTH immunonegative tumours (P = 0.9). Two of the patients with silent corticotroph adenomas have suffered multiple recurrences; one patient has had three operations and two courses of radiotherapy for two episodes of recurrence and one patient has had four operations, two courses of radiotherapy and gamma knife therapy after three recurrences in total. In contrast, no patient with an ACTH immunonegative tumour has required more than one course of treatment for tumour regrowth. CONCLUSIONS: This is the first single-centre comparative study of a series of clinically silent ACTH immunopositive tumours and has demonstrated that although they do not recur more often than ACTH immunonegative tumours, when they do regrow they show a more aggressive course. The practical implication of this is that there is no evidence for different postoperative imaging and radiotherapy protocols for ACTH immunopositive and immunonegative NFAs at initial presentation. However, if regrowth of a silent corticotroph tumour does occur then very careful monitoring is essential, after further treatment.