The use of high-dose corticosteroids in the treatment of severe sepsis and septic shock remains controversial. Our study was designed as a prospective, randomized, double-blind, placebo-controlled trial of high-dose methylprednisolone sodium succinate for severe sepsis and septic shock. Diagnosis was based on the clinical suspicion of infection plus the presence of fever or hypothermia (rectal temperature greater than 38.3 degrees C [101 degrees F] or less than 35.6 degrees C [96 degrees F]), tachypnea (greater than 20 breaths per minute), tachycardia (greater than 90 beats per minute), and the presence of one of the following indications of organ dysfunction: a change in mental status, hypoxemia, elevated lactate levels, or oliguria. Three hundred eighty-two patients were enrolled. Treatment--either methylprednisolone sodium succinate (30 mg per kilogram of body weight) or placebo--was given in four infusions, starting within two hours of diagnosis. No significant differences were found in the prevention of shock, the reversal of shock, or overall mortality. In the subgroup of patients with elevated serum creatinine levels (greater than 2 mg per deciliter) at enrollment, mortality at 14 days was significantly increased among those receiving methylprednisolone (46 of 78 [59 percent] vs. 17 of 58 [29 percent] among those receiving placebo; P less than 0.01). Among patients treated with methylprednisolone, significantly more deaths were related to secondary infection. We conclude that the use of high-dose corticosteroids provides no benefit in the treatment of severe sepsis and septic shock.