Objectives: Agriculture entered the discussions about safety climates late, despite being one of the most hazardous industries. We recently developed a safety climate scale in Bordeaux vineyards, for which we provided good evidence of reliability and construct validity (). In this study, we aimed to establish the external validity of this safety climate scale with the help of an independent national sample of vineyard professionals.Population and methods: We approached vineyard managers and operators during compulsory training and certification procedures for pesticide-related activities. Trainees giving informed consent for participation in the study were asked to complete a safety climate questionnaire at the start and end of a training session. In total, 406 vineyard managers or operators completed the questionnaire at the start of the study, 37 of whom declined to complete the questionnaire at the end of the training session, leaving 369 subjects available for pretraining/post-training comparisons. Statistical comparisons were based on t-tests and mixed models for repeated measures.Results: A mean safety climate score of 82.91 (SD: 9.06) was obtained in the initial survey in the Bordeaux region, whereas the safety climate score was estimated at 83.78 (SD 10.39) in this nationwide survey (P = 0.23). A significant increase was observed after the training course, for the safety climate score (7.5%, P < 10(-15)) and for each of its 7 dimensions (P < 10(-4) or less), in both univariate and multivariate analyses. However, the magnitude of these increases varied with dimension, ranging from 2.4% for rules and best practices to 15.5% for communication and feedback.Conclusions: External validity was demonstrated by transferability and sensitivity to intervention. This safety climate scale can now be considered to provide a good inference of the safety culture, with a meaning generalizable across vineyards.