The use of steroids is not required in myeloid malignancies and remains controversial in patients with acute lung injury (ALI) or acute respiratory distress syndrome (ARDS). We sought to evaluate dexamethasone in patients with ALI/ARDS caused by acute monocytic leukaemia (AML FAB-M5) via either leukostasis or leukaemic infiltration. Dexamethasone (10 mg every 6 h until neutropenia) was added to chemotherapy and intensive care unit (ICU) management in 20 consecutive patients between 2005 and 2008, whose data were compared with those from 20 historical controls (1994-2002). ICU mortality was the primary criterion. We also compared respiratory deterioration rates, need for ventilation and nosocomial infections. 17 (85%) patients had hyperleukocytosis, 19 (95%) had leukaemic masses, and all 20 had severe pancytopenia. All patients presented with respiratory symptoms and pulmonary infiltrates prior to AML FAB-M5 diagnosis. Compared with historical controls, dexamethasone-treated patients had a significantly lower ICU mortality rate (20% versus 50%; p = 0.04) and a trend for less respiratory deterioration (50% versus 80%; p = 0.07). There were no significant increases in the rates of infections with dexamethasone. In conclusion, in patients with ALI/ARDS related to AML FAB-M5, adding dexamethasone to conventional chemotherapy seemed effective and safe. These results warrant a controlled trial of dexamethasone versus placebo in AML FAB-M5 patients with noninfectious pulmonary infiltrates.