A young woman with fever and pleural effusion eventually was found to have both primary tuberculosis and systemic lupus erythematosus. Before either diagnosis was confirmed, the patient had a dramatic defervescence in response to prednisone; her condition had not improved initially after anti-tuberculous therapy, Uncertainty regarding the cli n ica I significance of this remarkable temperature response led us to review our experience with febrile tuberculous patients who had received corticosteroids. Of 14 such patients, nine had rapid declines of body temperature to below 36.56. In a comparable group of 14 patients with a varietyof connective tissue diseases, 12 showed an equally dramatic defervescence after corticosteroid administration. The data suggest that a prompt temperature decline following corticosteroid therapy is a nonspecific response that cannot be used to differentiate between fever of infectious or noninfectious origin.