OBJECTIVE To assess the evidence for adjunctive corticosteroids for severe community-acquired pneumonia (CAP). METHODS MEDLINE (1966-February 2007) and EMBASE (1980-February 2007) were searched to identify English- and French-language publications that evaluated the use of corticosteroids for CAP in adults. Major search terms included community-acquired pneumonia, intensive care unit, steroids, glucocorticoids, and adrenal cortex hormones. METHODS Clinical studies that evaluated the use of corticosteroids for CAP in adults were included. Clinical and surrogate markers of pneumonia were evaluated. RESULTS Severe CAP is associated with an increase in pulmonary and circulatory cytokines such as interleukin-6 and tumor necrosis factor-alpha that may be associated with higher mortality. Corticosteroids suppress inflammatory reactions and prevent migration of inflammatory cells from the circulation to tissues by suppressing the synthesis of chemokines and cytokines. One observational comparative study and 2 randomized, controlled studies examined the effects of corticosteroid therapy at various doses on endpoints of pulmonary and systemic inflammation and clinical outcomes. One small observational pilot study revealed that methylprednisolone blunted some of the pulmonary and systemic markers of inflammation. One small, randomized, placebo-controlled study revealed that hydrocortisone had no significant effects on markers of pulmonary and systemic inflammation or clinical outcomes. Another small, randomized, placebo-controlled preliminary study with methodological limitations revealed improvements in oxygenation, organ dysfunction score, and markers of inflammation favoring hydrocortisone over placebo. CONCLUSIONS Given the lack of proven benefit on clinically meaningful endpoints and adverse events, corticosteroids cannot be recommended for adjunctive treatment of severe CAP.