Purpose <p>This systematic review aims to identify independent etiological, anatomical, and procedural predictors of recurrent haemoptysis following Bronchial artery embolisation (BAE) using a Synthesis Without Meta-analysis (SWiM) framework.</p> Materials and Methods <p>PubMed, Cochrane, and Scopus were systematically searched in February 2026 for cohort studies analysing post-BAE haemoptysis recurrence in line with PRISMA guidelines. Data were synthesised using a SWiM approach as clinical heterogeneity and intra-cohort mixing of embolic agents precluded a meta-analysis. Multivariable evidence was prioritised over univariable results to determine the independent direction-of-effect, using a SWiM vote-counting approach. GRADE framework was used to assess the certainty of evidence.</p> Results <p>Sixty-five studies involving 12,143 patients met the inclusion criteria. Across multivariable evidence, systemic-pulmonary shunts, non-bronchial systemic collaterals (NBSCs), pleural thickening, bronchiectasis, aspergillosis, prior history of haemoptysis, positive sputum culture, and abnormal platelet indices were identified as positive independent predictors of recurrence. Conversely, active tuberculosis, pulmonary malignancy, post-tuberculous sequelae, high-volume haemoptysis, cystic fibrosis, and concurrent anticoagulant use were neutral predictors. Analysis of embolic agents did not demonstrate a clear superiority of any specific agent.</p> Conclusion <p>Chronic parenchymal destruction and systemic-pulmonary shunts are primary factors in long-term BAE failure due to haemoptysis recurrence as compared to presentation characteristics or embolic agents. Thorough angiographic investigation for cryptic collaterals and post-procedural management of underlying diseases is important for the efficacy of BAE.</p> Graphical Abstract <p></p>