RATIONALE: Lung injury caused by a ventilator results from nonphysiologic lung stress (transpulmonary pressure) and strain (inflated volume to functional residual capacity ratio). OBJECTIVES: To determine whether plateau pressure and tidal volume are adequate surrogates for stress and strain, and to quantify the stress to strain relationship in patients and control subjects. METHODS: Nineteen postsurgical healthy patients (group 1), 11 patients with medical diseases (group 2), 26 patients with acute lung injury (group 3), and 24 patients with acute respiratory distress syndrome (group 4) underwent a positive end-expiratory pressure (PEEP) trial (5 and 15 cm H2O) with 6, 8, 10, and 12 ml/kg tidal volume. MEASUREMENTS AND MAIN RESULTS: Plateau airway pressure, lung and chest wall elastances, and lung stress and strain significantly increased from groups 1 to 4 and with increasing PEEP and tidal volume. Within each group, a given applied airway pressure produced largely variable stress due to the variability of the lung elastance to respiratory system elastance ratio (range, 0.33-0.95). Analogously, for the same applied tidal volume, the strain variability within subgroups was remarkable, due to the functional residual capacity variability. Therefore, low or high tidal volume, such as 6 and 12 ml/kg, respectively, could produce similar stress and strain in a remarkable fraction of patients in each subgroup. In contrast, the stress to strain ratio-that is, specific lung elastance-was similar throughout the subgroups (13.4 +/- 3.4, 12.6 +/- 3.0, 14.4 +/- 3.6, and 13.5 +/- 4.1 cm H2O for groups 1 through 4, respectively; P = 0.58) and did not change with PEEP and tidal volume. CONCLUSIONS: Plateau pressure and tidal volume are inadequate surrogates for lung stress and strain. Clinical trial registered with www.clinicaltrials.gov (NCT 00143468).