Oxygen therapy is currently used as a supportive treatment in septic patients to improve tissue oxygenation. However, oxygen can exert deleterious effects on the inflammatory response triggered by infection. We postulated that the use of high oxygen concentrations may be partially responsible for the worsening of sepsis-induced multiple system organ dysfunction in an experimental clinically relevant model of sepsis. We used Sprague-Dawley rats. Sepsis was induced by cecal ligation and puncture. Sham-septic controls (n = 16) and septic animals (n = 32) were randomly assigned to four groups and placed in a sealed Plexiglas cage continuously flushed for 24 h with medical air (group 1), 40% oxygen (group 2), 60% oxygen (group 3), or 100% oxygen (group 4). We examined the effects of these oxygen concentrations on the spread of infection in blood, urine, peritoneal fluid, bronchoalveolar lavage, and meninges; serum levels of inflammatory biomarkers and reactive oxygen species production; and hematological parameters in all experimental groups. In cecal ligation and puncture animals, the use of higher oxygen concentrations was associated with a greater number of infected biological samples (P < 0.0001), higher serum levels of interleukin-6 (P < 0.0001), interleukin-10 (P = 0.033), and tumor necrosis factor-α (P = 0.034), a marked decrease in platelet counts (P < 0.001), and a marked elevation of reactive oxygen species serum levels (P = 0.0006) after 24 h of oxygen exposure. Oxygen therapy greatly influences the progression and clinical manifestation of multiple system organ dysfunction in experimental sepsis. If these results are extrapolated to humans, they suggest that oxygen therapy should be carefully managed in septic patients to minimize its deleterious effects.