Exercise is known to induce an immediate leucocytosis, the magnitude of which is related, in most instances, to the intensity and duration of the work. On finishing exercise, however, the leucocyte count may change in any one of several different ways. The pattern of postexercise changes in the leucocyte count is determined mainly by the time which has elapsed since beginning exercise, rather than the work intensity or the total work done, if, for example, exercise has been intermittent. Consideration of, firstly, the circumstances under which the plasma concentrations of catecholamines and cortisol have been found separately to correlate with the leucocyte count at the finish of exercise, and, secondly, the effects on the leucocyte count of exogenous administration of these substances has led us to develop a model which can satisfactorily account for all of the principal changes in the leucocyte count that have been noted during and after exercise. It is proposed that catecholamines produced during exercise act to increase the ratio of circulating to non-circulating leucocytes, while cortisol acts, by a mechanism which involves a time lag, to increase the total number of leucocytes in the vascular compartment. Examination of previously published reports shows that many contain results which support this model. Using the model as a basis, some predictions are made that can be tested experimentally, and some experiments are suggested which should help elucidate the mode of action of catecholamines and cortisol.