It is of great importance to be able quantitatively to evaluate the stress to which captive fur animals are subjected to by various forms of management. A range of haematological and hormonal parameters have been shown to be good indicators of stress. However, blood sampling, which entails the presence of humans and handling of the animals, can itself be a stressor that influences the results of an experiment. The present study shows that a group of silver foxes (n= 7) that had been regularly blood sampled (RBS-group) for a year had higher plasma concentrations of cortisol and lower concentrations of testosterone compared with an earlier non-blood-sampled group (NBS-group) (n= 7). The RBS-group also tended to have a lower number of lymphocytes. In addition, during a series of five blood samples taken at 30 min intervals, both groups showed an increase in plasma concentrations of cortisol, whereas numbers of lymphocytes and eosinophil leucocytes decreased. Plasma testosterone concentrations decreased after an initial increase. The study demonstrated that blood sampling, in both the long and short term, can profoundly affect hormone concentrations and leucocyte numbers.