The risk evaluation related to the occupational exposure to hazardous substances can be performed either by analytical data coming from environmental and/or biological monitoring or by using models and mathematical algorithms. Models are often considered economically more advantageous, nevertheless, in order to assure their reliability, they need to be validated by objective monitoring data. A predictive model for the evaluation of risks due to the exposure to pesticides in agriculture was elaborated, and, for model validations, 20 tomato farms were examined. For each farm, information necessary to the risk evaluation by using the elaborated algorithm were collected and, simultaneously, an environmental monitoring for the evaluation of inhaling and cutaneous exposure was carried out. The potential risk was estimated by comparing analytical data from environmental monitoring with respect to ADI (Admissible Daily Intake) values, for each investigated pesticide. Then, the accordance between the potential risk and the risk given by the elaborated model was evaluated and an agreement of 85.6% was found, showing a substantial reliability of the proposed model.