Coxiella burnetii infection occurs mainly after inhalation of aerosols contaminated by shedder ruminants. The propagation of C. burnetii between ruminant herds may result from a contiguity transmission between neighbours herds and/or the introduction of infected shedder animals in free herds. The objective of this study was to quantify and compare the relative impacts of the local cattle density and the introduction of possibly infected animals on C. burnetii spread between herds. Dairy herds located in France (n=2,829) and Sweden (n=117) were included. Herds were tested using ELISA kits in the bulk tank milk in May 2012 and June 2011 respectively. The total number of cattle introduced (ID) in herds was calculated during a period of 2 years before the date of ELISA test. The local cattle density was calculated using the total number of cattle located in a 5 km radius circle around herds. Logistic regression was used to assess the risk for a herd to be tested positive associated with an increase in the local cattle density and in the ID. The risk for a herd to be tested positive was higher for herds with a higher local cattle density (e.g. OR=4.05, CI95% 1.5-10.9, for herds with a local density between 100 and 120 compared to herds with a local density below 20). The risk was higher for herds with higher ID (e.g. OR=2.4, CI95% 1.7-3.4, for herds with ID over 3 compared to herds with no animal introduction). The variable country was not significant, indicating that results are not country specific. Population attributable fraction was 64% for the local density and 16% for ID. Results highlight the main role of contiguity in the dispersion process of C. burnetii between herds, suggesting the main role of wind. These results will enlighten the type of control measures to be implemented in herds at the regional level.