BACKGROUND: In patients with acute lung injury (ALI)/acute respiratory distress syndrome (ARDS), the use of alveolar-recruitment maneuvers to improve oxygenation is controversial. There is lack of standardization and lack of clinical studies to compare various recruitment maneuvers. Recruitment maneuvers are closely linked to the selection of positive end-expiratory pressure (PEEP), which is also a subject of debate. METHODS: With 12 intubated and mechanically ventilated patients with early ALI/ARDS we conducted a recruitment maneuver (sustained inflation at 40 cm H 2 O for 30 s), then set PEEP at 24 cm H 2 O, and then we reduced PEEP stepwise, by 4 cm H 2 O every 10 min. We kept the fraction of inspired oxygen (F IO 2) at 0.8. After each PEEP decrement step we measured P aO 2. We defined the "optimal" PEEP as the PEEP step above which P aO 2 decreased by > 20%. All the patients then underwent a period of ventilation on the same settings: tidal volume 6 mL/kg, PEEP at the level set by the physician before the experiment, plateau pressure < 30 cm H 2 O. Then each patient underwent 3 ventilation strategies, each applied for one hour: optimal PEEP alone; optimal PEEP plus one sustained inflation (40 cm H 2 O for 30 s); and optimal PEEP plus sigh breaths (ie, twice the baseline tidal volume, plateau pressure < 40 cm H 2 O) every 25 breaths. After the application of each PEEP strategy we measured arterial blood gas values and the static compliance of the respiratory system. RESULTS: The mean ؎ SD optimal PEEP was 12 ؎ 4 cm H 2 O. The measurements from the standardization periods were comparable between the 3 PEEP groups. In the optimal-PEEP-plus-sighs group the changes in P aO 2 (85 ؎ 96%) and static compliance (14 ؎ 20%) were significantly greater than in the 2 other groups. CONCLUSIONS: Sighs superimposed on lung-protective mechanical ventilation with optimal PEEP improved oxygenation and static compliance in patients with early ALI/ARDS.