Characterizing climate and climate change impacts on species in mountainous and other complex terrain can be complicated by microclimatic variation. Most studies rely on widely available free-air climatic models like PRISM and ClimateNA —models of choice for assessing how terrain-related differences in climate exposure affect ecological responses at various scales. However, the accuracy of these models in complex terrain and their relationship to the microclimates pikas experience remains poorly understood. At the Niwot Ridge Long-Term Ecological Research site, we installed 16 stations measuring talus free-air, surface and subsurface temperatures hourly from summer 2022 to summer 2024 to: 1) evaluate whether ClimateNA accurately predicts free-air temperature, 2) investigate how free-air temperatures relate to talus surfaces and subsurface temperatures year-round, and 3) develop statistical models that allow us to incorporate microclimatic temperature into assessments of pika occupancy. Our results highlight systematic biases in ClimateNA predictions that could substantially mischaracterize thermal stress for thermally sensitive species like the American pika, a species with only 3°C separating normal and lethal body temperatures. These findings underscore the necessity of developing free-air temperature correction models and recognize that even within free-air environments, complex terrain creates buffered microclimates critical for thermally sensitive indicator species.