Pesticide exposure increases the risk of chronic disease among farmers. Understanding exposure is necessary for epidemiological and regulatory purposes. In Europe, worker exposure is assessed during the registration process using the OPEX model, which is based on a limited number of studies, often unpublished and carried out by pesticide companies. We assessed the conservativeness of OPEX for workers performing postapplication tasks (re-entry, harvesting). Methods: In 2016–2017, dermal exposure to captan/THPI and dithianon was measured in French fruit farm workers during 65 re-entry (net folding and deployment, thinning, tying) and 58 harvesting days, using patches and cotton gloves. We used linear regression to compare measured and corresponding OPEX-calculated exposure using 1) default parameters; 2) field parameters (actual task duration, measured dislodgeable foliar residues) for 20 observations. Results: Workers were exposed several days after the last application, which is not considered in the pesticide registration process. We found that the model underestimated exposure calculated with field parameters in all observations for dithianon and 60% for captan, linked to an underestimation of OPEX transfer coefficients (ratio of 0.40 for captan and 0.26 for dithianon between default and measured transfer coefficients). Discussion: When observation occurred several days after application, OPEX tended to underestimate exposure. An industry study conducted under controlled working conditions found divergent results. It seems important to include field studies conducted under usual working conditions in the registration process to ensure a truly conservative approach and to consider cumulative exposure, since post-application tasks account for around 600 working hours a year.< Réduire