The aim of this study was to evaluate the safety and diagnostic yield of bedside bronchoalveolar lavage (BAL) combined with fibrescopic transbronchial lung biopsy (TBLB) in determining the aetiology of pulmonary infiltrates in mechanically ventilated patients.The records of 38 mechanically ventilated patients who underwent BAL/TBLB to investigate unexplained pulmonary infiltrates were retrospectively reviewed. Patients were divided into two groups: immunocompetent (group 1: n=22; group 1a: n=11, late acute respiratory distress syndrome (ARDS); group 1b: n=11, no ARDS) and immunocompromised (group 2, n=16).The procedure allowed a diagnosis in 28 patients (74%), inducing therapeutic modification in 24 (63%) and confirmation of clinical diagnosis in four (11%). In groups 1a, 1b and 2, diagnosis was obtained in 11 out of 11 (fibroproliferation), seven out of 11 and 10 out of 16 patients, and therapy changed in 11 out of 11 (administration of steroids), six out of 11 and seven out of 16 patients, respectively. Pneumothorax occurred in nine patients (four of group 1a), bleeding in four (<35 mL), and transient hypotension in two. No fatalities were procedure-related.Combined bronchoalveolar lavage/transbronchial lung biopsy is of diagnostic and therapeutic value in mechanically ventilated patients with unexplained pulmonary infiltrates, excluding those with late acute respiratory distress syndrome. Although complications are to be expected, the benefits of the procedure appear to exceed the risks in patients in whom a histological diagnosis is deemed necessary.