Introduction Studies suggest an increased risk of lung cancer in the hard metal industry. This issue is under study in an international cohort study. An Austrian plant participates in the study. Here we focus on four questions: (1) Can we estimate cumulative exposures among cohort members from routine industrial health (IH) data? (2) Are the IH data supported by urine data? (3) What is the relation between cobalt in the air and in urine? (4) Can we estimate temporal trends in exposure so that gaps in IH measurements in earlier years can be closed?

Methods Work histories were collected from all workers hired since 1.1.1950. IH data of dust, tungsten, and cobalt in air and of cobalt in urine were obtained. Air data were log-transformed and entered into multiple logistic regression models as dependent variables. Log-Levels of dust, cobalt and tungsten were explained by a linear time trend, by analytical method, and department. This regression model allowed predicting exposures at other times as well. Predicted exposures of cobalt were entered as explanatory variable for urine values in a second regression model.

Results Overall, work histories were collected from 1969 workers. Repeated urine tests were available from 253 workers from the years 2008–2014. Air measurements of dust (n = 130), tungsten (n = 141) and cobalt (n = 147) exist for the years 1985 to 2012. For all substances a significant decrease over the years is detectable. The air and urine values of cobalt correlate with each other: One mg/m³ cobalt in the air leads to an excretion of 200 µg/l cobalt in urine.

Conclusion The available data allow an initial assessment of the inhalation exposure of these workers cohort. The Austrian occupational limit value for cobalt (0.5 mg/m³) is too lenient compared to the limit in urine (10 µg/l).