Introduction: Diagnosis of pulmonary infiltrates in immunocompromised hosts (ICH) remains challenging. Haematology patients may be at risk of complications from bronchoscopy due to cytopenias. Aims and Objectives: A prospective registry of ICH referred for bronchoscopy was established in a tertiary university-affiliated hospital to provide data on patient outcomes (CIRB No:2021/2148). Results: We recruited 77 ICH over 2 years. 42 (54.5%) had haematological diagnoses: leukaemia 16 (38.1%), lymphoma 10 (23.8%), myelodysplastic syndrome 8 (19.1%), multiple myeloma 4 (9.5%), aplastic anaemia 2 (4.8%), myelofibrosis 1 (2.4%), amyloidosis 1 (2.4%). Non-haematological diagnoses: HIV 4 (5.2%), renal transplant 5 (6.5%), solid organ cancer 19 (24.7%), autoimmune disease 14 (18.2%). Haematology patients were more likely to have received cytoreductive chemotherapies (p=0.014), less likely to have received steroids 2 weeks before bronchoscopy (p=0.002); had lower FiO2 requirements at time of referral (0.25 vs 0.40, p=0.002), haemoglobin levels (p<0.001), platelet counts (p<0.001) and moderate neutropenia ANC < 1 (p=0.001). We found no significant differences in bronchoalveolar lavage yield of gram stain and culture, AFB smear and culture, significant galactomannan antigen level, fungal culture, and respiratory viral PCR. Compared to non-haematology ICH, risks of bleeding requiring bronchoscopic measures (3 vs 0, p=0.11), intubation (2 vs 3, p=0.01) and critical care admission within 24 hours of bronchoscopy (3 vs 3, p=0.05) were not increased. Conclusion: In our centre, adverse events from bronchoscopy were not increased in haematology patients.