The pathogenesis of corticotroph adenomas is unknown. In a recent study accumulation of p53 protein was detected by immunohistochemistry in a substantial proportion of pituitary corticotroph adenomas, and it has been suggested that it may be causally related to their development. However, other immunohistochemical studies have not confirmed the high incidence of p53 accumulation in this tumor type. Therefore, in the present study, p53 protein accumulation was re-examined in a series of 31 cases of corticotroph adenomas, using different sets of well validated anti-p53 antibodies. Furthermore, in view of the known association of p53 protein with apoptosis, and the known property of p53 to form complexes with heat shock proteins (HSPs), the relationship of p53 accumulation in corticotroph adenomas with apoptosis and HSP-70 was also investigated. Tumor samples from 31 patients with Cushing's disease or Nelson's syndrome were studied. Accumulation of p53 protein was tested by the standard ABC method using two different sets of clone Pab1801 and DO-7 monoclonal antibodies, applied after incubation of sections in a microwave oven. Using the DO-7 antibody, nuclear accumulation of p53 protein was detected in a total of 15 cases, with cytoplasmic staining observed in only 3 tumors. In contrast, using the Pab1801 antibody nuclear staining was observed in only 5 adenomas, with 11 adenomas demonstrating focal cytoplasmic immunoreactivity. Parallel sections of all corticotroph tumors demonstrating cytoplasmic accumulation of p53 protein were tested for the immunohistochemical presence of heat shock protein HSP-70. A striking similar distribution pattern of these two proteins was observed. Apoptosis, identified by the in situ end labeling technique, was detected in a total of 15 out of 28 corticotroph adenomas tested. Calculation of the apoptotic labeling index (ALI) by image analysis showed a significantly lower ALI in those corticotroph adenomas demonstrating nuclear p53 accumulation compared to those with no nuclear p53 immunostaining (p < 0.05). There was no significant difference in the ALI between cytoplasmic p53 positive and negative tumors. It is concluded that depending on the antibody used there is a significant variation of p53 protein detection in corticotroph adenomas. Overall, a significant proportion of corticotroph adenomas studied expressed the p53 protein, which depending on the antibody used, was located either in the nucleus and/or the cytoplasm of tumorous corticotroph cells. Cytoplasmic accumulation of p53, as shown by our colocalization studies with HSP-70, may be due to p53/HSP-70 complex formation. Although such a complex-mediated cytoplasmic exclusion of p53 has no significant effect on apoptosis, nuclear accumulation of p53 protein is associated with a significantly lower apoptotic index indicating a failure of p53 protein to exert its apoptotic action in at least a subset of this tumor type.