This thesis records the findings of a retrospective study of decompression illness (DCI) in the UK compressed air tunnelling industry since the mid-1980s. The thesis describes how the study arose, its scope and objectives, along with an overview of tunnelling and shaft-sinking. The development of compressed air working techniques is reviewed along with a description of decompression practice and DCI, and an outline of relevant legislation and guidance. The acquisition and manipulation of data to form a number of databases and spreadsheets on which the analysis was performed is discussed. That analysis examined measures of DCI incidence and quantified that incidence using these measures. Also considered is the variation in tolerance and susceptibility to DCI in the workforce, and the phenomenon of acclimatisation. An examination of the extent to which men worked on multiple contracts and the variation in their susceptibility to DCI on these contracts is included. The main conclusions are that a number of measures of DCI incidence were identified, some more appropriate than others and that the incidence of DCI when so measured was high, disproportionately so in shift workers. No reasonably practicable restrictions on exposure were identified which would have allowed the retention of air-only decompression. Oxygen decompression looked promising but had yet to be used sufficiently extensively to generate enough data for analysis. Recommendations included one that an alternative technique for monitoring the effectiveness of decompression should be developed. The thesis ends with recommendations for further research.