This article discusses the oxygen consumption (VO2) and delivery (DO2) relationship as it pertains to animal models of sepsis and human sepsis syndrome and septic shock. Pathologic dependence of VO2 on DO2 is not present in resuscitated patients who have sepsis syndrome and septic shock. Defects in oxygen extraction and use at the individual organ level with maldistribution of blood flow probably do occur in sepsis; however, there is no clinical evidence that augmenting DO2 to supernormal levels decreases organ dysfunction or mortality in sepsis. We need improved techniques to assess tissue hypoxia at the organ level, and we need to test therapies directed at correcting the maldistribution of blood flow and O2 use defects of sepsis.