Emerging or re-emerging diseases involving both livestock and wildlife represent a prominent challenge to disease management. Indeed, knowledge on the hosts or the disease, as well as surveillance data, are often scarce in such cases. This effectively limit our ability to understand transmission dynamics and predict the efficacy of management strategies. However, when enough data or prior knowledge is available, mechanistic modelling can help us to better understand disease transmission within multi-host systems. In particular, it can disentangle the contribution of the different host populations to the maintenance and the transmission functions. The maintenance function is defined as the capacity to sustain the infectious agent within the system. In contrast, the transmission function is defined as the capacity to transmit to the target population, e.g. domestic livestock. A reservoir can then be defined as one or several host populations that fulfil both the maintenance and the transmission functions. The aim of our study was precisely to better understand which host populations contribute to the maintenance and transmission functions in the case of brucellosis in the French Alps. This multi-host system includes Alpine ibex (Capra ibex) and Alpine chamois (Rupicapra rupicapra), as well as domestic ruminants (cattle, sheep and goats). Evidence of cross-contamination was found in ibex, chamois and cattle, with very close genetic distance between the isolated Brucella melitensis strains. We developed a multi-host, compartmental, deterministic model of Brucella transmission within and between these different host populations. Taking advantage of surveillance data in both wild and domestic species, we estimated within- and between-population transmission rates. Contrary to previous studies in animal health, where contact patterns were hypothesized, we also integrated results from observational studies on direct and indirect contacts between species to inform our estimations. From the estimates of the transmission rates, we were then able to calculate the within- and between-population reproduction numbers, defined as the expected number of new infections in a host population generated by a single infectious individual in the same or different host population. By calculating which population or sets of populations had a reproduction number above one (the threshold for infection persistence), our results allowed to bring insights on the populations’ contribution to the maintenance function. Similarly, we identified the populations that contributed to transmission to domestic cattle. These results will inform surveillance and management of brucellosis in this multi-host system.