Objective To study the patterns of plasma concentrations of endotoxin, tumor necrosis factor-α (TNF), interleukin-6 (IL-6), plasminogen activator inhibitor-1, C-reactive protein, and serum amyloid A during the treatment of human sepsis. Design A prospective case series study. Setting ICU of the Department of Internal Medicine, University Hospital Groningen, The Netherlands. Patients Twenty consecutive patients (11 female, 9 male, mean age 67 yrs) with clinically defined sepsis. Eighteen patients were admitted from the outpatient emergency ward; two patients were already inpatients. The control group (n = 7) comprised patients with nonseptic shock. Measurements and Main Results Ten (50%) septic patients had detectable endotoxemia (>5 ng/L). TNF concentrations on admission were increased in 94% of the septic patients, whereas IL-6 and plasminogen activator inhibitor plasma concentrations were increased in all septic patients. The septic group showed significantly (p < .05) higher concentrations of TNF, IL-6, plasminogen activator inhibitor-1, C-reactive protein, and serum amyloid A compared with the nonseptic patients. In the septic group, we found a correlation of both IL-6 and plasminogen activator inhibitor concentrations with severity of illness (r2 = .33, p < .05; r2 = .22, p < .05, respectively). After the start of antibiotic treatment, high concentrations of TNF and plasminogen activator inhibitor persisted in the nonsurvivors in contrast to decreasing concentrations in most of the survivors. After an initial increase in seven patients, IL-6 concentrations decreased in all septic patients and also in nonsurvivors. Conclusions This study confirms previous findings that: a) TNF is a major mediator involved in the pathogenesis of septic shock; b) plasminogen activator inhibitor activity is significantly increased in septic patients and might be involved in the pathogenesis of disseminated intravascular coagulation associated with sepsis; and c) IL-6 is involved in the pathophysiology of septic shock, although further studies are needed to determine whether IL-6 is directly involved in mediating the lethal complications of septic shock or whether it should be considered an “alarm hormone” that reflects endothelial cell injury. Our findings also suggest that the concentrations and trends of these mediators during treatment are valuable for monitoring septic patients.