The epidemiological link between brucellosis in wildlife and brucellosis in livestock and people is widely recognised. When studying brucellosis in wildlife, three questions arise: (i) Is this the result of a spillover from livestock or a sustainable infection in one or more host species of wildlife? (ii) Does wildlife brucellosis represent a reservoir of Brucella strains for livestock? (iii) Is it of zoonotic concern? Despite their different host preferences, B. abortus and B. suis have been isolated from a variety of wildlife species, whereas B. melitensis is rarely reported in wildlife. The pathogenesis of Brucella spp. in wildlife reservoirs is not yet fully defined. The prevalence of brucellosis in some wildlife species is very low and thus the behaviour of individual animals, and interactions between wildlife and livestock, may be the most important drivers for transmission. Since signs of the disease are non-pathognomonic, definitive diagnosis depends on laboratory testing, including indirect tests that can be applied to blood or milk, as well as direct tests (classical bacteriology and methods based on the polymerase chain reaction [PCR]). However, serological tests cannot determine which Brucella species has induced anti-Brucella antibodies in the host. Only the isolation of Brucella spp. (or specific DNA detection by PCR) allows a definitive diagnosis, using classical or molecular techniques to identify and type specific strains. There is as yet no brucellosis vaccine that demonstrates satisfactory safety and efficacy in wildlife. Therefore, controlling brucellosis in wildlife should be based on good management practices. At present, transmission of Brucella spp. from wildlife to humans seems to be linked to the butchering of meat and dressing of infected wild or feral pig carcasses in thedeveloped world, and infected African buffalo in the developing world. In the Arctic, the traditional consumption of raw bone marrow and the internal organs of freshly killed caribou or reindeer is an important risk factor.