Twenty years have now elapsed since Ashbaugh and Petty first described the syndrome of acute respiratory failure associated with a wide spectrum of clinical conditions. During the past two decades, significant advances have emerged in our understanding of the clinical conditions associated with the syndrome and the pathophysiological changes affecting the alveolar-capillary membrane responsible for the characteristic non-cardiogenic pulmonary edema. Recent data have reaffirmed the notion that mortality rates in ARDS remain in excess of 60 percent, essentially unchanged since the first description of the syndrome, despite all the advances in critical care medicine in the intervening years. The incidence of ARDS has been difficult to establish because of lack of agreement on precise definition criteria. The lack of agreed definition criteria has hampered evaluation of the natural history of the syndrome, its epidemiology and mortality rates, and the efficacy or otherwise of a variety of therapeutic interventions. This review will highlight a recent, clinically appropriate, expanded definition of ARDS. New understandings of the roles of sepsis and multi-system organ failure in mortality associated with ARDS will be discussed. Several mediators, both locally in the lung and in the systemic circulation, have been implicated in the pathophysiology of ARDS. This review will discuss the evidence for and against neutrophils, platelets, cytokines derived from mononuclear cells and macrophages, complement, prostaglandins/leukotrienes, oxygen-derived radicals, and a variety of proteases. Current treatment strategies for ARDS are designed to increase tissue oxygen delivery by increasing arterial oxygen tension and cardiac output while simultaneously attenuating the pulmonary and systemic injury by appropriate pharmacologic and surgical interventions. Recent data advocating pharmacological augmentation of cardiac index and oxygen delivery will be highlighted. The persistently high mortality rates of 60-70 percent in patients with established ARDS have provoked recurring interest in new techniques of providing mechanical ventilation. Most studies have shown, however, that mortality in ARDS patients is attributable mainly to sepsis and multi-system organ failure rather than primarily to respiratory failure. Established and speculative intervention to reduce sepsis and multi-system organ failure associated with ARDS will be featured in the review.