Processes related to the spatiotemporal spread of infections in metapopulations are most often partially observed, especially for long-lasting endemic diseases. Bovine paratuberculosis (agent Mycobacterium avium subsp. paratuberculosis – Map), is a worldwide enzootic disease of economic importance for dairy cattle producers. Its screening in the field is difficult due to the long incubation period and the low sensitivity of diagnostic tests. Our objective was to estimate key parameters of a multiscale dynamical model of Map spread from longitudinal disease-related data, and thus to provide new insights on its infection dynamics. We developed a mechanistic model of Map spread between dairy herds, accounting for stochastic within-herd demography and infection dynamics, and animal trade between farms. We modelled 2,030 dairy herds in Brittany (France), for which comprehensive data on cattle trade and partial data on animal infection status were available (2005-2013). Data was augmented by back-tracing locations of animals with known health statuses. Using Approximate Bayesian Computation and Sequential Monte Carlo, we estimated the probability of purchasing infected cattle from outside the metapopulation (p), indirect local transmission rate (tr), and sensitivity of diagnostic test (Se). The numerical identifiability was also investigated. Our results suggested that p was between 0.3 and 0.7, in agreement with current knowledge on Map prevalence at herd level. The value of tr was close to the assumed value, whereas Se seemed systematically biased towards low values. Our model allows parameter estimation from various data sources, accounting for observational errors, to provide accurate predictions of infection dynamics