The usefulness of data on absence from work due to sickness (absenteeism) as an indicator of specific and detailed information of diagnoses is widely discussed in the international literature. In occupational health indicators for health risks (e g sickness absence of more than 14 days), details on diagnosis and workplace are very useful, if analysed by epidemiological means. A pilot project, initiated by a high sickness absence in the automobile industry, led to a test of an industrial health information system, which was abruptly stopped by data protection arguments. The epidemiologic approach of an age-adjusted comparison of incidences relies on the access to morbidity data of the health insurance system. With the diagnostic information at hand differences of incidences in one production branch (as recognized by the incidence ratio SIR) are to be discovered if related to certain specific diagnoses. Thus, a direct comparison of comparable rates helps to identify specific reasons for increased absence from work. Practical solutions were found if the owner of the data, the National Federation of Enterprise Sickness Funds, conducted the analyses. First results showed as yet unknown associations of myocardial infarction and obstructive lung diseases in the metallurgical industry. In such cases not only the validity of such results has to be verified, but also other epidemiological tools, such as a case-referent approach to determine risk ratios, are required for the identification of any causally important relationship.