Assisted colonisation – the introduction of a species outside its indigenous range for conservation purposes – has been suggested for several decades as a pragmatic response to climate change. But virtually no one has tried it. In this presentation I describe a collaboration between researchers and managers where assisted colonisation is being trialled to reduce the risk of extinction of a Critically Endangered freshwater turtle. The western swamp turtle is a long-lived reptile native to Perth and occupies seasonal wetlands on which it depends for food and reproduction. Much of its former habitat has been cleared, and the small fragments that remain are slowly drying due to declining rainfall and groundwater. Conservation translocations of juveniles raised in captivity have generally been successful, with more than 1000 individuals now released, but the long-term climatic suitability of translocation sites is a key concern. Consequently, based on biophysical modelling, multiple criteria analysis and empirical data, we have shifted our focus to evaluating release sites in cooler and wetter parts of south-west Australia. Beginning in 2016, juvenile turtles have been released into novel wetlands across a 350 km latitudinal gradient to study their growth, physiological performance, and survival. Now into our fourth trial, we have used each translocation experiment to answer increasingly complex questions, and to focus on the most promising wetlands for assisted colonisation south of the species known range. Our evidence-based approach is potentially a much-needed global example to break the apparent inertia between policy making and implementation of assisted colonisation initiatives.