1. Non-invasive faecal cortisol metabolite (FCM) analysis is a well-established tool to quantify stress in captive and free-ranging species. While the method has great potential, its suitability in field studies might be limited when faecal samples from unknown individuals are used. Possible factors affecting final results and thus jeopardizing correct data interpretation are individual and sex-specific variation, storage conditions and uneven distribution of metabolites in the faeces. 2. We tested these factors on a population of free-ranging Alpine chamois Rupicapra rupicapra rupicapra in the Austrian Alps. Faecal samples (n = 183) were analysed with an established enzyme immunoassay (EIA). To further validate the assay for FCM in chamois, a high-performance liquid chromatography (HPLC) was performed. Sex-specific differences in metabolite excretion were evaluated. Effects of storage length and temperature on FCM were tested with two experiments. The distribution of metabolites in the faeces was determined by the analysis of subsamples of single faecal samples. Potential individual effects on FCM levels and individually variable reactions to stressful events were evaluated with a simulation experiment. 3. Patterns of immunoreactive peaks after HPLC separation were similar for different faecal samples, except in one sample of a male. In the stability tests, storage time at ambient temperature prior to freezing and the individual were the most important variables in modulating FCM. Concentrations within single samples varied significantly between pellets. Analysis of faecal samples collected from June to October showed a highly significant seasonal trend (P < 0.001) and a considerable variance of FCM levels within the population. Simulations confirmed that individual reactions to stressors in terms of varying gradients and FCM levels can explain the observed FCM patterns. 4. Using FCM to assess adrenocortical function requires measuring extensively metabolized products of glucocorticoids, whose excretion and detection in faeces depend on several environmental, endogenous and methodological factors. In free-ranging wildlife, these factors and the intrinsic individual differences in FCM excretion generate systemic noise and substantially distort final results. Therefore, sampling of unknown individuals inevitably jeopardizes meaningful interpretation of data, if the above named factors are not taken into consideration.