In order to examine the feasibility and safety of undertaking a larger prospective study to compare the diagnostic yield from concurrent open lung biopsy (OLB) and bronchoalveolar lavage (BAL) in febrile neutropenic patients with pulmonary infiltrates and the impact of such knowledge on clinical outcome, a pilot exploratory study was performed. 13 immunocompromised patients (mainly with haematological malignancy or bone marrow transplantation recipients) were investigated. At least one diagnostic finding in 12 of 13 patients was provided by OLB compared to 4 of 13 patients by BAL. BAL provided 7 specific diagnoses (pneumocystis 1, fungal infection 3, bacterial pneumonia 1, pulmonary haemorrhage 2) whilst OLB provided 12 specific diagnosis (CMV 2, pneumocystis 3, fungal infection 1, bacterial pneumonia 1, pulmonary haemorrhage 4, pulmonary embolism 1). Five patients with nonspecific interstitial/alveolar inflammation were diagnosed only by OLB. The concordance that the exact same specific diagnoses present in the OLB were found in the BAL was zero. There were 2 minor complications (1 wound infection by OLB, 1 moderate haemorrhage by BAL). Mortality at 28 days was 8 of 13 patients which in no case was related to either procedure. We suggest that OLB is a safe procedure in such patients, provides superior and more complete diagnostic information compared to BAL and a larger controlled study to investigate the impact of early OLB on the outcome of these patients appears to be justified.