Context: To re-establish a population of the threatened numbat (Myrmecobius fasciatus) in a newly created safe haven at Mount Gibson Wildlife Sanctuary, in the semi-arid zone of Western Australia, wild-born and captive-bred individuals of both sexes were translocated. Aim: To compare: (1) the use of refuges by wild-born numbats pre- and post-translocation, and (2) the survival and refuge use of captive-bred numbats compared with wild-sourced numbats post-translocation. Methods: Monitoring via radio-tracking facilitated the gathering of information on survival and behaviour of numbats pre- and post-translocation, and allowed a comparison of how different factors (e.g. captive-bred or wild-born; male or female) influenced survival and establishment in a new environment. Key results: There were no significant differences in survival between sexes or between wild-born and captive-bred individuals. However, there were some differences in behaviour between sexes and source populations. Captive-bred numbats, regardless of sex, made greater use of tree hollows as nocturnal refuges than did their wild-born counterparts. Among wild-born numbats, there was a comparatively greater use of tree hollows and logs on the ground by males at Mount Gibson than at Scotia. The use of diurnal escape refuges did not vary between sexes or between captive-bred and wild-born individuals. Conclusions: On the basis of the information presented here, we conclude that, in the absence of predation by mammalian predators, and with suitable release habitat, captive-bred-to-wild translocations of numbats may be as likely to succeed as are wild-to-wild translocations, at least over the first few months post-translocation. Implications: Optimising the size and genetic diversity of the founding population by using animals from a variety of sources may not need to be heavily constrained by concerns about the ability of captive-bred numbats to adapt to and survive life in the wild.