Many epidemiological studies of the relation between work and disease use information on work history obtained by interview from the study subjects. A validation study was undertaken to evaluate the accuracy of this information collected from 100 workers in a shipbuilding industry. The information furnished by the workers was compared with that present in the company's registers. The work history (job titles and starting dates) was relatively accurate and the validity varied with the number of events to declare and with their duration; it also depended on the type of information and the precision required.