U.S. national parks protect a natural heritage of global significance; those parks, especially those in the arid southwest, are threatened by climate change. Identifying climate refugia within our national parks using not only statistical models, but also validating predictions using robust field data should provide focus for managers in their stewardship of parks' biological resources. In the region surrounding Joshua Tree National Park (JTNP), which straddles the Colorado and Mojave deserts in southern California, previous research has predicted the widespread demise of its namesake iconic species, the Joshua tree (Yucca brevifolia) due to climate change. In order to assess whether climate refugia exist for Joshua trees in the future at JTNP, we employed both field measurements and statistical models. We used current distribution point data together with historic climate data, to match conditions when the existing Joshua trees established, in order to predict the distribution of continuously suitable conditions (refugia) at the end-of-century. While the high and moderate mitigation could result in refugia for approximately 19% and 14% of the original area within JTNP, respectively, the business-as-usual scenario indicated an almost complete elimination of Joshua trees from the park. In order to validate model predictions, using teams of community scientists, we measured the demographic patterns of Joshua tree stands from low to upper elevations within JTNP. Recruitment within stands shows a strong concordance with modeled climate refugia; high-recruiting stands were within or closer to modeled refugia and in areas with lower climatic water deficit, higher precipitation, and lower maximum temperature than low-recruiting stands. These findings most importantly indicate the importance of regional to global mitigation strategies for carbon emissions, as reflected in the difference between maintenance of refugia vs. an almost complete elimination of the species from the park by the end-of-century. This also underscores the need to protect areas predicted to support refugia from multiple management threats. Rather than an ominous prediction of extinction, climate refugia provide land stewards with targets for focusing protective management, giving desert biodiversity places to weather the future.