Rats are widely used in biomedical research as animal models for human diseases. However, due to their small body size, blood sampling is complicated and invasive and thereby can seriously interfere with endocrine functions and possibly compromise the animals' welfare. Therefore, a non-invasive technique to monitor stress hormones in these animals is highly desired. Our study aimed to gain general information about corticosterone metabolism and excretion and to validate a 5alpha-pregnane-3beta,11beta,21-triol-20-one enzyme immunoassay (EIA) to reliably measure faecal corticosterone metabolites (CMs) in laboratory rats. In total, 18 rats were administered 2.3 MBq of (3)H-corticosterone intravenously and per os, respectively (intravenous: 6 males and 6 females; per os: 3 males and 3 females). Subsequently, all voided excreta were frequently collected for five days. About 75+/-9% of the recovered CMs were found in the faeces. Peak concentrations of radiolabelled steroids appeared in the urine after 1.7+/-0.6 h in males and after 6.0+/-3.5 h in females. In faeces, maxima were observed after 14.7+/-2.4 h in both sexes. In principle, the time course and delay for both routes of administration (intravenous or per os) were the same, except for a delay of peak concentrations in urine (4.5+/-2.1 h) in per os administered males. Using high-performance liquid chromatography (HPLC), faecal (3)H-CMs were characterized and differences were found between the sexes. In both sexes, corticosterone was extensively metabolized, but while males showed only minor variations in their CM patterns, those of females differed largely between individuals. To validate the mentioned EIA, we investigated the diurnal variation (DV) of glucocorticoids as well as effects of the injection procedure itself and conducted an adrenocorticotropic hormone challenge test and a dexamethasone suppression test, using six male and six female rats each. Our results demonstrated that pharmacological stimulation, suppression and DV of adrenocortical activity were accurately reflected by means of CM measurement in faeces. By successful physiological validation, we proved for the first time the suitability of an immunoassay to non-invasively monitor adrenocortical activity in rats of both sexes. This method opens up new perspectives for biomedical and pharmacological investigations as well as for animal welfare related issues.