Acute respiratory distress syndrome (ARDS) was described more than 50 years ago. 1 In that first description, Ashbaugh et al 1 framed ARDS as an acute lung injury that arises from several underlying causes, such as severe trauma or pneumonia, and can lead to catastrophic hypoxemia and death. The authors considered several potential therapies that might modulate the underlying disease process, but concluded the only demonstrably useful intervention was optimal titration of positive end-expiratory pressure (PEEP). They acknowledged, however, that PEEP worked principally by improving oxygenation as a temporizing measure while waiting for the underlying cause of ARDS to resolve. ARDS rarely occurs in isolation, and the precipitating causes often contribute not only to respiratory failure, but also to multisystem organ failure. Thus, with advances in mechanical ventilation, such as better use of PEEP, refractory hypoxemia is now rarely the cause of death for patients with ARDS. 2 In recent decades, however, it has become increasingly apparent that lifesaving mechanical ventilation is, paradoxically, injurious in its own right, causing ventilator-induced lung injury (VILI). VILI is caused by pressure-and volumemediated trauma to the airways and alveoli from ventilatorgenerated breaths, inducing a systemic inflammatory response that in turn contributes to multiple organ failure and death. 3 The recognition of the potentially harmful effects of mechanical ventilation has shifted the goal of optimal PEEP away from just improving oxygenation toward preventing VILI. 4 "Optimizing" PEEP is not easy. PEEP splints the lung open at the end of expiration. It can therefore minimize atelectrauma by decreasing cyclic opening and closing of lung units, but can also lead to overdistention. 3 The effect of PEEP on driving pressure, an important determinant of VILI, therefore depends on the balance of its effects on recruitment of nonaerated areas vs overdistention of normally aerated ones. 5 In addition, higher PEEP can impede venous return, compromising right ventricular function and central hemodynamics. In other words, some level of PEEP is needed to maintain oxygenation and to avoid VILI, but how much? And which is the best method to find it? Several randomized trials have compared higher vs lower levels of PEEP in patients with ARDS. [6][7][8][9] Individual trials did not show a beneficial effect of higher levels of PEEP on mortality, although an individual patient data meta-analysis suggested higher PEEP reduced mortality in patients with ARDS