The primary aim of surgery in the management of cholesteatoma is eradication of the disease which can potentially result in serious complications such as intracranial extension, facial nerve weakness and further hearing loss. A profound hearing loss resulting postoperatively considerably compromises the outcome but is sometimes unpredicted. In this series the authors looked at 617 middle ear operations performed by a single consultant in which there were six cases of ‘dead ear’. None of these followed 83 operations for otosclerosis or 62 paediatric mastoid operations. However in 187 adults undergoing exploration of the mastoid for cholesteatoma, five (2.7% of cases) ended up with ‘dead ears’. The author has described these six cases in which the pathology ranged from fistulae of the lateral semicircular canal, granulations over the footplate, cholesteatoma covering the footplate, cholesteatoma surrounding the ossicular chain and unanticipated extensive cholesteatoma. In two of these cases surgery went satisfactorily and ‘dead ear’ was not at all predictable, but in the other four further hearing loss appeared likely in the course of surgery. It appears that in one of these cases ‘dead ear’ was definitely avoidable, in two it was probably avoidable and in the other three total hearing loss was inevitable in the process of clearance of disease. All the cases had CT scans which helped to some extent to decide on the surgical approach, but did not reliably predict hearing outcome after surgery. In this series 2% of patients had a ‘dead ear’ due to cholesteatoma itself. Therefore, the risks involved with surgery can be weighed against the disease itself causing a ‘dead ear’. This requires careful thought on whether or not to operate and its timing. The incidence of 2.7% ‘dead ear’ following cholesteatoma surgery is rather high and this calls for informed consent with very clear explanation of surgery and the possible adverse outcomes.