Previous reports have shown that endotoxin decreases serum iron in experimental animals. In this study fever was produced in nine female and nine male normal subjects in order to define the temporal and quantitative changes in serum iron and ferritin concentrations. Six volunteers were randomly given bacterial endotoxin (5 ng/kg) or saline intravenously and received the alternative compound a week later. Serial blood samples were drawn at 4-hr intervals for a 24-hr period, beginning when the compound was administered, for the determination of serum iron and ferritin concentrations. The same study was performed with intramuscular etiocholanolone (0.3 mg/kg) or the vehicle, propylene glycol, as a control, but the first blood sample was obtained 9 hr after the compound was given. In addition, blood samples were obtained at 12-hr intervals in six volunteers for 11 days after an intramuscular injection of etiocholanolone. The results showed a significant increase (p less than 0.005 for etiocholanolone, P less than 0.01 for endotoxin) in serum ferritin and a significant decrease (p less than 0.005 for etiocholanolone, p less than 0.001 for endotoxin) in serum iron for both pyrogenic compounds compared with the control compounds. However, the amount of fever and the changes in the iron parameters were greater with etiocholanolone. One episode of induced fever with etiocholanolone effected changes in serum ferritin and iron concentrations that lasted 10 days. Thus this study demonstrated that a single episode of fever in man produced rapid and prolonged changes in serum iron and ferritin concentrations.