Although fever is a well known and common manifestation of hypersensitivity, little has been learned about the sequence of events by which antigen disturbs temperature regulation in sensitized individuals. Using tuberculininduced fever in rabbits infected with BCG as a model, we have shown in previous papers that an endogenous pyrogen (EP) appears in the circulation of such animals during the febrile response (1, 2). When injected intravenously this agent causes immediate fevers in normal rabbits, and is, therefore, clearly distinct from tuberculin itself which produces fever after a latency of 45 to 60 minutes and is pyrogenic only in animals previously sensitized by BCG. The tissue source of EP in this form of hypersensitivity fever has remained uncertain. In several systems of hypersensitivity, including tuberculin, Johanovsk~ has reported that when antigen is added in vitro to mononuclear ceils from spleen or lymph nodes of specifically sensitized animals, a pyrogen is formed which he has called 'hypersensitivity pyrogen' to distinguish it from serum endogenous pyrogen present in other experimental fevers (3-8).In the following experiments, tuberculin was added in vitro to various tissues of BCG-sensitized rabbits in an attempt to repeat Johanovskg's findings and to determine the origin of the pyrogen which appears in the serum in tuberculin-induced fever. Johanovsk~'s observations were not confirmed, and evidence is presented instead that sensitized blood cells release EP in vitro when incubated with tuberculin and hence may be a source of the circulating pyrogen in this form of hypersensitivity fever.In additional experiments, it was found that normal blood cells, briefly incubated in plasma of sensitized animals, were similarly activated in vitro by tuberculin, an observation which suggests that humoral antibodies play a role in the genesis of tuberculin-induced fever.