Completeness of company records is a crucial point in interpretation of follow-up studies of industrial cohorts. The Employers Quarterly Reports on Earnings (EQRE) are valuable for control of company records in the United States. The public Supplementary Pension Scheme (ATP) and the industrial statistics questionnaires are of equal value in Denmark. The authors report on a cohort study from a chemical plant, where cohort members were identified from company records, which were checked against ATP and industrial statistics data. The mortality analysis shows employees known only from the ATP data to have an excess mortality, relative risk (RR) = 1.45. Inclusion in the cohort of these additional employees only changed the relative risk in overall mortality from 1.01 to 1.04, but turned out to be decisive in the study of a rare disease.