Because invasive predators pose a threat to Australian mammals, we are increasingly preserving threatened populations within predator-free havens. Havened populations are commonly used as source populations for translocation, but many translocations of havened populations have failed. In the absence of predators inside havens, havened mammals can lose their anti-predator traits, and it is predicted that individuals with such weakened anti-predator responses should have reduced survival and reproduction when faced with predators, though few empirical tests of this theory exist. We conducted an experimental release whereby 40 havened and 40 non-havened woylies (Bettongia penicillata ogilbyi; Australia’s most translocated species) were translocated to an area of vacant bushland where they are exposed to a low density of foxes and feral cats. For one year, we have monitored fitness and behaviour among these two cohorts from which we will draw comparisons between the predator naïve havened population and the predator exposed non-havened population, describing any adaptive changes in anti-predator responses over time. We will also compare rates of survival and reproduction in the havened and non-havened populations to evaluate the suitability of the havened population for future translocations. These findings can directly inform translocations of this and other populations of havened mammal to ensure the best survival outcomes. Our study paves the way toward a conceptual and mechanistic understanding of the loss of anti-predator responses, which is essential if we are to aim for the persistence of havened species outside of predator-free havens.