The United Kingdom Atomic Energy Authority mortality study investigated the relation between mortality and recorded exposure to ionising radiation among employees working at the authority's seven establishments between 1946 and 1979. This report examines the design of the study and methods of data collection and validation. The completeness of the study population was deemed to be unsatisfactory at two establishments, where records of employment before 1965 had been destroyed. Assessment of the magnitude of the deficit led to the conclusion that the data from these establishments were too incomplete for inclusion in the mortality analysis. At the other establishments validation showed that the data collected were accurate and unbiased. Certain characteristics of the 39 546 employees included in the mortality analysis were identified which were relevant in interpreting the findings.