The aim of this study was to estimate occupational exposure to inhalable wood dust by country, industry, the level of exposure and type of wood dust in 25 member states of the European Union (EU-25) for the purposes of hazard control, exposure surveillance and assessment of health risks. National labour force statistics, a country questionnaire (in 15 member states, EU-15), a company survey (in Finland, France, Germany and Spain), exposure measurements (from Denmark, Finland, France, Germany, The Netherlands and the United Kingdom) and expert judgements were used to generate preliminary estimates of exposure to different types of wood dust. The estimates were generated according to industrial class (six wood industries, four other sectors) and level of exposure (five classes). These estimates were reviewed and finalized by national experts from 15 member states. Crude estimates were generated also for 10 new member states (EU-10). The basic data and final estimates were included in the WOODEX database. In 2000-2003, about 3.6 million workers (2.0% of the employed EU-25 population) were occupationally exposed to inhalable wood dust. Of those, construction employed 1.2 million exposed workers (33%), mostly construction carpenters. The numbers of exposed workers were 700,000 (20%) in the furniture industry, 300,000 (9%) in the manufacture of builders' carpentry, 200,000 (5%) in sawmilling, 150,000 (4%) in forestry and <100,000 in other wood industries. In addition, there were 700,000 exposed workers (20%) in miscellaneous industries employing carpenters, joiners and other woodworkers. The numbers of exposed workers varied by country ranging from <3,000 in Luxembourg and Malta to 700,000 in Germany. The highest exposure levels were estimated to occur in the construction sector and furniture industry. Due to limited exposure data there was considerable uncertainty in the estimates concerning construction woodworkers. About 560,000 workers (16% of the exposed) may be exposed to a level exceeding 5 mg m(-3). Mixed exposure to more than one species of wood and dust from wooden boards was very common, but reliable data on exposure to different species of wood could not be retrieved. This kind of assessment procedure integrating measurement data, company data, country-specific data and expert judgement could also serve as one model for the assessment of other occupational exposures.