Introduction Potentially preventable deaths in police custody include those which involve illicit drugs, alcohol and deliberate self-harm. Near miss incidents (NMI) that did not result in death have a crucial role in understanding risk factors in custody. Such research has not previously been undertaken. A program of research has been developed to study NMI, in order to better identify those at risk in police custody. For the purposes of this research, NMI have been defined as ‘an unplanned and unforeseeable or unforeseen event that could have resulted, but did not result, in human death or may have resulted in injury or other adverse outcomes’. It was intended that the definition although broad, would not include simple accidents (e.g. slipping on urine in a cell) or trivial injury. Aims and methods The two aims of the study are (a) to determine whether it is realistic to attempt to assess NMI with the intention of identifying information of use in enhancing detainee care and (b) to assess how frequently NMIs occur and whether there are specific patterns. Pilot interviews were conducted with three forensic physicians practising in London, UK to create a structured questionnaire for all forensic physicians working in London. The questionnaire provided the basis of a retrospective recall survey of all forensic physicians working in London as Forensic Medical Examiners. The questionnaire was designed to assess the numbers of NMI, patterns in occurrence and relevant learning points within the previous 6 months. A covering letter, background questionnaire (exploring the background of the medical practitioner), copies of the survey, and reply paid envelopes were sent to each Forensic Medical Examiner ( n = 134) in London, contracted to provide forensic medical services for the Metropolitan Police Service. Data about all incidents were anonymised. Results Ninety six (73%) Forensic Medical Examiners responded. Of these 18% were Principal grade, the remainder were Senior (24%), Standard (35%) and Assistant (23%). Thirty eight NMI were reported by 27 Forensic Medical Examiners (of all levels). The initial reason for police contact was recorded as alcohol ( n = 8), theft and robbery ( n = 7), warrants ( n = 4), violence ( n = 3), traffic violations ( n = 2) and single cases of drugs, murder and immigration offences. Of the main perceived cause of each NMI, illicit drugs were involved in 12/38, alcohol in 17/38, deliberate self-harm in 11/38, issues concerning searches, checks or rousing in 8/38, failure of inter-agency communications in 5/38, and possible resource issues in 4/38. In a number of cases more than one factor was involved. Conclusions The information established about NMI is broadly consistent with documented patterns of deaths in police custody in England and Wales which supports the validity of the data. The next stage of this research will be a prospective six month study in which NMI will be analysed in order to learn lessons which may be utilised to attempt to prevent potentially avoidable deaths in police custody.