Introduction: There is a major challenge in determining pesticide exposure among agricultural workers. This is due to the wide variety of plant protection products (PPPs) used on a single crop and the range of active ingredients used over time. Materials and Methods: A study was conducted in 2010 among a sample of workers covered by the Agricultural Social Security Scheme in five French departments (Coset-MSA pilot study) and collected, via a self-administered questionnaire, data on current employment. The concept of straw cereal crops was cross-referenced with a specific crop-exposure matrix (CEM), developed as part of the Matphyto program. This program involves the creation of crop-exposure matrices for PPPs, comprising, for each crop, indices characterizing exposure to PPPs by major group, by major chemical family, and by specific active ingredient. The probabilities of exposure to PPPs in 2010 were estimated by summing the individual exposure probabilities provided by the CEM for the various groups and families of pesticides used in France on straw cereals. Partial non-response to the questionnaire was accounted for by imputation using administrative data available in the Agricultural Scheme’s information systems, and total non-response by reweighting. Results: Initial estimates show that in 2010, 21.1% (95% CI: 19.2–23.1) of active ingredients covered by the Agricultural Scheme in the five departments considered were used on farms where straw cereals were treated with fungicides, 20.1% (95% CI: 18.2–21.9) with herbicides, and 15.6% (95% CI: 14.2–17.0) with insecticides. Triazoles were the most commonly used fungicides, pyrethroids the most common insecticides, and as for herbicides, numerous chemical classes were used on straw cereals, with no single class dominating. Conclusion: The work carried out successfully tested the cross-referencing of data collected during the pilot phase of the Coset-MSA cohort with one of the MCE databases from the Matphyto program and verified that the collected data are sufficient for this cross-referencing. This analysis also demonstrates the value of administrative data in understanding partial non-response to the questionnaire. This work will be continued by taking into account the entire career and other crops. Its application to the national Coset-MSA cohort will provide population-level indicators for the agricultural scheme and identify the populations most occupationally exposed to PPPs in France.