Background: Climate change refugia, areas buffered from climate change relative to their surroundings, are of increasing interest as natural resource managers seek to prioritize climate adaptation actions. However, evidence that refugia buffer the effects of anthropogenic climate change is largely missing. Methods: Focusing on the climate-sensitive Belding's ground squirrel (Urocitellus beldingi), we predicted that highly connected Sierra Nevada meadows that had warmed less or shown less precipitation change over the last century would have greater population persistence, as measured by short-term occupancy, fewer extirpations over the twentieth century, and long-term persistence measured through genetic diversity. Results: Across California, U. beldingi were more likely to persist over the last century in meadows with high connectivity that were defined as refugial based on a suite of temperature and precipitation factors. In Yosemite National Park, highly connected refugial meadows were more likely to be occupied by U. beldingi. More broadly, populations inhabiting Sierra Nevada meadows with colder mean winter temperatures had higher values of allelic richness at microsatellite loci, consistent with higher population persistence in temperature-buffered sites. Furthermore, both allelic richness and gene flow were higher in meadows that had higher landscape connectivity, indicating the importance of metapopulation processes. Conversely, anthropogenic refugia, sites where populations appeared to persist due to food or water supplementation, had lower connectivity, genetic diversity, and gene flow, and thus might act as ecological traps. This study provides evidence that validates the climate change refugia concept in a contemporary context and illustrates how to integrate field observations and genetic analyses to test the effectiveness of climate change refugia and connectivity. Conclusions: Climate change refugia will be important for conserving populations as well as genetic diversity and evolutionary potential. Our study shows that in-depth modeling paired with rigorous fieldwork can identify functioning climate change refugia for conservation.