Virtually of the all recent therapeutic interventions for treating sepsis have failed to improve survival. One potential explanation is that the heterogeneity of the immune response to the septic challenge is such that only a portion of the patients die as a result of excessive inflammation. The clinical trials lacked power because traditional measurements do not accurately identify these patients. Previous work has shown that higher levels of interleukin (IL)-6 are found in those mice that die from septic peritonitis; therefore, we sought to determine whether IL-6 measured 6 h after surgery could predict outcome. Adult, female BALB/c mice (n = 79) were subjected to cecal ligation and puncture with a 21-gauge needle and treated with imipenem in D5W every 12 h for 5 days, resulting in a homogenous population at the outset. Six hours after surgery, 20 microL of blood was obtained from the tail vein to measure IL-6. Mortality was followed for 21 days. Overall 3-day survival was 77%, and 21-day mortality was 56%. Plasma IL-6 levels >2,000 pg/mL were determined to predict mortality within the first 3 days with a sensitivity of 58% and specificity of 97%. To further refine the mortality prediction, body weight and a complete blood count were performed 24 hours after cecal ligation and puncture. Discriminate analysis indicated that a weighted formula combining body mass, lymphocyte, and platelet count would predict death with sensitivity of 83% and a specificity of 79%. We tested the value of the IL-6 prediction by surgically resecting the cecum in those animals with IL-6 > 2000 pg/mL, which resulted in a significant improvement in survival. These data demonstrate that IL-6 measured 6 h after injury accurately predicts mortality resulting from experimental sepsis. This measurement may be determined quickly so that therapy may be targeted only to those individuals at significant risk of dying and initiated within sufficient time to be effective.