Urban forests provide ecosystem services to more than 4.2 billion people living in cities; however, the provision of these services is threatened by climate change. Cities will experience novel, warmer climates that will impact tree species survival. Here, we present an empirical approach combining the use of climate analogues and vulnerability metrics to guide urban tree species selection. Climate analogues can be used to identify where the current climate in a given location is similar to the projected future climate of another location. Using vulnerability metrics, tree species in both locations can then be compared and species at risk of future climate change can be identified and substituted with species known to be resilient to climatic changes in the other location. Using Canadian cities as a case study, we show how our approach can become a useful tool to inform species selection and facilitate urban forest decision-making in response to climate change.