In many regions of sub Saharan Africa large mammals occur in human-dominated areas, yet their community composition and species-specific densities have rarely been described in areas occupied by traditional hunter-gatherer and pastoralist groups. Surveys of the mammal populations in such areas are also provide important context to understand human hunting practices, including the choice of prey species harvested and the overall productivity of hunting. Using a sampling grid centered on a Hadza hunter-gatherer camp and covering 36 km² of semi-arid savannah in northern Tanzania, we assessed mammals using camera traps (n = 19 stations) for c. 5 months (2,182 trap nights). Overall, we recorded 36 wild mammal species and rarefaction curves at the site- and landscape scale suggested that sampling effort was sufficient to capture mammal species richness. Species-specific densities were estimated using a random encounter model and site- and speciesâ body mass- specific estimates of the area sampled at each camera; confidence intervals were estimated using bootstrapping. Point estimates of densities varied by c. four orders of magnitude, from 0.003 ind./km² (African wild dog) to 27.5 ind./km² (Kirkâs dik dik). Densities of livestock (cattle, donkey, sheep and goat) were high, particularly when estimated using directly observed herd sizes. The biomass density of livestock exceeded that of all wild mammals by a factor of 3.3-38.7. Compared to camera trap rates in a fully protected area of northern Tanzania (Lake Manyara National Park), most wildlife species in our study area appear in much lower abundance, and the magnitude of species-specific differences in relative abundances were not significantly associated with body mass. We outline and discuss specific hypotheses that could explain the observed patterns of high species richness and generally low species densities. Finally, we discuss how these data inform studies of Hadza hunting and models of hunter-gatherer diets.