BACKGROUND: In various countries, reports from occupational physicians (OPs) are an important source of work-related illness (WRI) data. In Great Britain (GB), this is undertaken through the Occupational Physicians Reporting Activity (OPRA) surveillance scheme. Because access to an OP is uneven in GB applying the GB workforce as the denominator could lead to bias when calculating incidence rates. To improve the validity and utility of OP-derived data, it is important to improve the quality of the underlying denominator data. AIMS: To estimate the proportion of the GB workforce covered by OPRA participants and subsequently to calculate OP-derived incidence rates. METHODS: OPRA participants were surveyed once in each triennial period (2005-07 and 2008-10) about the workforce they covered. Numbers of GB employees within each major industrial division covered by the OPRA reporters' occupational health (OH) services were calculated and compared with Labour Force Survey data. Incidence rates were calculated for all industry and for selected sectors. RESULTS: OPRA reporters' OH services covered ~13% of the GB workforce in 2005-07 and 9% in 2008-10. This increased to 34% if adjusted to represent all GB OPs. Annual average incidence rates (2005-07 and 2008-10) were 301 and 336 (total WRI), 150 and 199 (mental ill-health), 103 and 99 (musculoskeletal), 23 and 24 (skin), and 11 and 9 (respiratory), per 100 000 employed. CONCLUSIONS: Estimating the workforce covered by OP reporters can strengthen the quality of the information source, enabling comparisons between OP data and information from other sources, as shown by OPRA in GB.