Objectives In recognition of the significant impairment caused by haemoptysis on a patient's quality of life, bronchial artery embolisation has been introduced worldwide as one of the first-line treatment options. Since little evidence is available on the mechanisms of recurrent haemoptysis after super-selective bronchial artery coil embolisation (ssBACE), the purpose of the present study is to evaluate these. Methods We retrospectively evaluated the mechanisms of recurrent haemoptysis using both enhanced computed tomography and cineangiography following ssBACE by reviewing 299 haemoptysis-related arteries (HRAs) in 57 consecutive patients who underwent 2nd series ssBACE for the management of recurrent haemoptysis between April 2010 and December 2015.Results Median age of patients was 69 (interquartile range 64-74) years, and 43.9% were men. This study revealed that (1) recanalisation was the most common mechanism (45.2%) followed by development of new HRA (38.5%), bridging collaterals (14.7%) and conventional collaterals (1.7%); (2) these trends could be modified in several situations such as with antiplatelet or anticoagulant medications; (3) relatively large-diameter HRAs were more likely to recanalise compared with small-diameter HRAs and (4) recurrent haemoptysis could be managed by 2nd series ssBACE with a procedural success rate of 97.7% without any major complications. Conclusions Recanalisation was the most common mechanism of recurrent haemoptysis after ssBACE. Our results provide interventionists with indispensable insights. Key Points • Recanalisation was the most common mechanism of recurrent haemoptysis after super-selective bronchial artery coil embolisation, followed by development of new haemoptysis-related arteries • These trends could be modified in several situations such as with antiplatelet or anticoagulant medications • Recurrent haemoptysis could be managed by 2nd series super-selective bronchial artery coil embolisation with a procedural success rate of 97.7% without any major complications.