Objectives Population-based incidence rate data is essential when assessing employment sectors at risk. This study aimed to investigate the best methods of estimating the size and characteristics of a population derived from General Practitioners (GPs) from which incidence of work-related ill-health was determined. Methods The Health & Occupational Reporting network in General Practice (THOR-GP) collects information on work-related ill-health from 250 to 300 GPs. The THOR-GP denominator is the registered population of participating GPs9 practices. Two methods of estimating the population were developed; one using practice postcode and another using the registered patients9 postcodes. Population data was broken down by occupational and industry. Results from both methods were compared with each other and national figures. Results Patient postcodes were received from 110 GPs totalling 940 401 postcodes. These were mapped to corresponding geographical output areas (OA). Each OA was weighted according to the number of patients and the total population broken down by occupation and industry. When compared to population data based on the corresponding practice postcodes the proportional distribution by industry was found to be almost identical, and very similar to the national picture (eg, Healthcare; practice postcode 11.6%, patient postcode 11.5%, GB 11.8%). Conclusions Estimating the denominator using the ‘gold standard’ of patient postcode data and comparing it to practice postcode information validates the use of practice postcode in the absence of individual patient information in GP based health studies. Although only a small proportion of GB GPs report to THOR-GP, results have shown that they are nationally representative.