Mobile species pose major challenges for conservation because of their unpredictable, large scale movements in response to fluctuating resources. If locations with critical resources overlap with threats, large proportions of a mobile population may be exposed to threats. Critically endangered and nomadic swift parrots Lathamus discolor nest wherever food is most abundant in their breeding range, but concern exists that nest predation from an introduced predator may severely affect their population. Although swift parrots nest on predator‐free offshore islands, population viability analysis indicates that is land nesting alone may be insufficient to offset extinction risk from high mainland predation rates assuming that the species is a single panmictic population. We test the assumption that swift parrots act as a single conservation unit. We undertook a population genetic analysis using seven microsatellite loci and samples obtained over 6 years from across the breeding range of swift parrots. We found no evidence of departure from Hardy–Weinberg expectations across the species and both Analysis of Molecular Variance and Bayesian Structure analyses failed to detect any evidence for genetic differentiation across the samples both spatially and temporally. These results, supported by simulations, indicate panmixia and a lack of population genetic structure in swift parrots. Unlike a sedentary or site philopatric species, the majority of the swift parrot population may be at risk of exposure to predation when unpredictable resources draw individuals away from islands. These findings support a key assumption of population viability models that predict an extreme reduction in population size for swift parrots, and address a major gap in knowledge of the species’ ecology. Our study has implications both for the development of effective conservation management strategies and for the longer‐term evolution of avoidance of predator‐infested habitat in swift parrots.