The main aim of this thesis was to try to assess the feasibility of identifying people at high risk of death from coronary heart disease by simple screening procedures. The approach to this problem was based upon the analysis of a series of surveys which had been organised in the West of Scotland. These included both general population surveys and industrial surveys and the oldest, which was a mixture of both types was started in 1965. Information about the deaths of anyone surveyed had been supplied via the Registrar General and was available up to August 1977. The first analysis of these data was to compare several different statistical methods of arriving at the probability of dying from coronary heart disease in a fixed length of time. The oldest survey was used for this comparison which appeared to show that of the methods tried the multiple logistic method was the most suitable one. After this a model was put forward which endeavoured to model the complete survival function. As time from survey was now incorporated in the model, it was possible to try out the same model on all the data from the different surveys, even though there was a difference in follow-up time from 2 years on one file to 12 years on the oldest. This met with a fair measure of success but the final questions about the forms of the model could not be resolved until another data file, this time of a survey of civil servants in London, was brought in. This led to one model being put forward which seemed to give an adequate fit on all the files that had complete data for every criterion of fit proposed. This model was based on males from 45 to 65 years old and though it seemed to extend to males under 45 it did not turn out to be at all appropriate for women, though the small amount of data on women meant that it was not feasible to discover whether modifications of the model could have led to a form suitable for women. However, the success of the model in providing a good fit to data from different places, classes and with very different follow-up periods gives real hope that it might provide a basis for a screening procedure to identify people at high risk of coronary heart disease.