An evidence-based approach to the conservation management of a species requires knowledge of that species’ status, distribution, ecology, and threats. Coupled with budgets for specific conservation strategies, this knowledge allows prioritisation of funding toward activities that maximise benefit for the species. However, many threatened species are poorly known, and determining which conservation strategies will achieve this is difficult. Such cases require approaches that allow decision-making under uncertainty. Here we used structured expert elicitation to estimate the likely benefit of potential management strategies for the Critically Endangered and, until recently, poorly known Night Parrot (Pezoporus occidentalis). Experts considered cat management the single most effective management strategy for the Night Parrot. However, a combination of protecting and actively managing existing intact Night Parrot habitat through management of grazing, controlling feral cats, and managing fire specifically to maintain Night Parrot habitat was thought to result in the greatest conservation gains. The most cost-effective strategies were thought to be fire management to maintain Night Parrot habitat, and intensive cat management using control methods that exploit local knowledge of cat movements and ecology. Protecting and restoring potentially suitable, but degraded, Night Parrot habitat was considered the least effective and least cost-effective strategy. These expert judgements provide an informed starting point for land managers implementing on-ground programs targeting the Night Parrot, and those developing policy aimed at the species’ longer-term conservation. As a set of hypotheses, they should be implemented, assessed, and improved within an adaptive management framework that also considers the likely co-benefits of these strategies for other species and ecosystems. The broader methodology is applicable to conservation planning for the management and conservation of other poorly known threatened species.