A dynamic spatial stochastic model of the spread of Coxiella burnetii between dairy cattle herds is presented. The objective of the study is to identify the dominant route of transmission for a major region for dairy cattle farming (Finistère) in France. First, the ability of the model to replicate the observed annual incidence (in number of herds) was assessed. Based on the receiver operating characteristics analysis, the model performed best at infection probability cut-off of 0.28. The model predicted 261 incident herds on average, with sensitivity of 76% and specificity of 77%. At this infection probability cut-off, 79% incident herds were infected because of airborne transmission of the pathogen, while the rest of the herds were infected because of introduction of an infectious cow during the cattle trade. Irrespective of the infection probability cut-off, windborne transmission caused the majority of the introductions compared to the transmission due to cow trade