In this issue of CHEST, we have two reports of the use of cytokines as markers of illness: Cardier and coworkers (see page 2238), who report on the levels of tumor necrosis factor (TNF)-α/TNF receptor I (TNFRI) and Fas ligand/Fas in sepsis; and a report by Vincent and coworkers (page 2232), who report on the TNF-2 allele and survival from cardiogenic shock. These articles illustrate how these markers may have a future in clinical practice much as routine laboratory values did at the start of the last century. Both articles showed how specific levels of cytokines can play a role in patient outcome, transforming growth factor-β1 being predictive of survival in cardiogenic shock, and increased levels of TNFRI and Fas as important markers of the severity of sepsis in the other. Both articles measured and evaluated TNF-α one of the most commonly measured cytokines in research. TNF-α has been implicated in myocardial depression associated with septic shock,4as well in the pathogenesis of heart failure. The study by Vincent et al showed that the TNF-2 allele of the TNF-α promoter is a strong independent factor associated with better survival from cardiogenic shock. The clinical implication cannot be understated, in that it if we can identify patients with cardiogenic shock with the TNF-2 allele we may be able to provide them with more aggressive therapy or transfer them to university centers were they may have improved outcomes. TNF-α levels in the study by Cardier et al, as in several other studies,5–6 found no correlation in mortality in patients with severe sepsis, which suggests that although TNF-α plays an important role in the pathogenesis of sepsis, it may not be a good marker or predictor of mortality in severe sepsis. TNF-α levels can also be greatly affected by the support provided in the ICU. Our group7 showed how there can be a loss of compartmentalization of TNF-α in the lung and can affect systemic levels based on the mode of mechanical ventilation. This varied utility of one common cytokine marker shows how specific cytokine levels can be used in specific patient populations and why further investigation of the utility of each cytokine must be investigated in critical illness.