Anthropogenic climate change is generating mismatches between the environmental conditions that populations historically experienced and those in which they reside. Understanding how climate change affects population performance is a critical scientific challenge. We combine a quantitative synthesis of field transplant experiments with a novel statistical approach based in evolutionary theory to quantify the effects of temperature and precipitation variability on population performance. We find that species’ average performance is affected by both temperature and precipitation, but populations show signs of local adaptation to temperature only. Contemporary responses to temperature are strongly shaped by the local climates under which populations evolved, resulting in performance declines when temperatures deviate from historic conditions. Adaptation to other local environmental factors is strong, but temperature deviations as small as 2°C erode the advantage that these non-climatic adaptations historically gave populations in their home sites.