BACKGROUND To determine whether plateau pressure (Pplat) measurement is lowered and peak airway pressure (Pawpeak)-to-Plat gradient is increased by measurement on a decelerating compared with square gas delivery wave form. METHODS Prospective before and after study of mechanically ventilated injured and critically ill patients in an adult surgical intensive care unit. Pplat, Pawpeak, and Pawpeak-to-Pplat gradient were measured on decelerating and square gas delivery wave forms. RESULTS Pplat and other routine ventilator parameters were measured in 82 (47 trauma, 35 emergency general surgery) consecutive convenience sampled adult intensive care unit patients on decelerating and then square gas delivery wave forms. Peak gas flow was fixed at 40 L/min; all other parameters (rate, tidal volume, positive end-expiratory pressure) were held constant. All patients were managed on assist control volume cycled ventilation using fentanyl and midazolam or propofol; no neuromuscular blockade was used. Patients with Pawpeak more than 35 cm H2O were excluded. Comparing decelerating with square gas delivery, mean Pawpeak was lower (25.1 ± 2.3 cm H2O vs. 33.1 ± 2.1 cm H2O; p < 0.0001) and mean Pplat was lower (21.3 ± 1.9 cm H2O vs. 24.8 ± 2.5 cm H2O; p < 0.0001), resulting in a decreased Pawpeak-to-Pplat gradient (3.8 ± 2.1 vs. 8.3 ± 2.3; p < 0.0001). CONCLUSIONS Changing from a decelerating to a square gas delivery wave form significantly increases Pplat and Pawpeak, thereby increasing the Pawpeak-to-Pplat gradient. This increase may prompt unwarranted therapy aimed at reducing the gradient to its normal value of 4 cm H2O pressure or less. Conversely, patients with a high Pawpeak on a square wave form may benefit from transitioning to a decelerating wave form before changing ventilation parameters. METHODS Diagnostic study, level III.