Introduction - In France, an estimated 4.5% to 8% of new cancer cases are believed to be linked to occupational exposure. It is currently not possible to calculate cancer incidence rates by occupation and industry sector. Using applications for long-term medical care (ALD) is one option to consider to address this gap. However, before implementing a supplementary collection of occupational data based on ALD applications, it was essential to determine their completeness rate, as well as the number of false positives. Method - A study conducted by the Francim network on the concordance between ALD data and registry data (gold standard) in five departments provides answers to this issue. The sensitivity and positive predictive value (PPV) of ALD data were examined for three sites of major interest in the field of occupational risks: lung, bladder, and leukemias. Results - Sensitivity was approximately 50% for the three sites studied. The PPV varied by site: 46% for the bladder, 63% for leukemias, and 83% for the lung. Conclusion - The low sensitivity of ALD does not support their use as the sole data source for producing cancer incidence rates by occupation and industry sector in France. Furthermore, it is likely that applications for ALD status are linked to socio-occupational factors. (R.A.)