Nonlinear temperature effects indicate severe damages to U.S. crop yields under climate change. This study examines the relationship between temperature and yields of three major U.S. crops: corn, soybeans, and cotton. Using a panel of county-level yields and a new fine-scale weather dataset, the research finds that yields increase with temperature up to specific thresholds (29°C for corn, 30°C for soybeans, and 32°C for cotton), after which they decline sharply. The relationship is nonlinear and asymmetric, with a steeper decline above the optimal temperature than below it. This pattern is consistent across time-series and cross-sectional analyses, suggesting limited historical adaptation of seed varieties or management practices to warmer temperatures. The study predicts that area-weighted average yields will decrease by 30–46% under the slowest warming scenario (B1) and 63–82% under the most rapid warming scenario (A1FI) by the end of the century. These predictions are based on the Hadley III climate model. The results highlight the significant impact of extreme high temperatures on crop yields, with the projected increase in frequency of such temperatures being a key driver of these large predicted impacts. The study also finds that the same nonlinear temperature effect is observed in both the cross-section of counties and the aggregate year-to-year time series. The relationship between yield and temperature is consistent across different regions of the U.S., including cooler northern states and warmer southern states. The temperature-yield relationship observed between 1950 and 1977 is the same as that observed between 1978 and 2005, despite technological advancements that have increased average yields. Greater precipitation partially mitigates damages from extreme high temperatures, but the study finds no significant correlation between temperature and precipitation outcomes in the raw daily data. The estimated climate-change impacts are insensitive to the specified growing season and consistent with time separability. The study combines fine-scale weather data with flexible regression models, finding consistent nonlinear temperature effects across time, locations, crops, and variations in temperature and precipitation. The study also notes that the estimated relationship is consistent when comparing estimates based on year-to-year weather variations and cross-sectional climate variations. This suggests limited historical adaptation to extreme heat for any given crop. The findings highlight the importance of considering nonlinear temperature effects in climate change models and the need for further research on potential adaptations to climate change, such as changing crop locations or seasons. The study also acknowledges the limitations of its approach, including the inability to account for CO₂ concentrations, which may affect plant growth and yields.Nonlinear temperature effects indicate severe damages to U.S. crop yields under climate change. This study examines the relationship between temperature and yields of three major U.S. crops: corn, soybeans, and cotton. Using a panel of county-level yields and a new fine-scale weather dataset, the research finds that yields increase with temperature up to specific thresholds (29°C for corn, 30°C for soybeans, and 32°C for cotton), after which they decline sharply. The relationship is nonlinear and asymmetric, with a steeper decline above the optimal temperature than below it. This pattern is consistent across time-series and cross-sectional analyses, suggesting limited historical adaptation of seed varieties or management practices to warmer temperatures. The study predicts that area-weighted average yields will decrease by 30–46% under the slowest warming scenario (B1) and 63–82% under the most rapid warming scenario (A1FI) by the end of the century. These predictions are based on the Hadley III climate model. The results highlight the significant impact of extreme high temperatures on crop yields, with the projected increase in frequency of such temperatures being a key driver of these large predicted impacts. The study also finds that the same nonlinear temperature effect is observed in both the cross-section of counties and the aggregate year-to-year time series. The relationship between yield and temperature is consistent across different regions of the U.S., including cooler northern states and warmer southern states. The temperature-yield relationship observed between 1950 and 1977 is the same as that observed between 1978 and 2005, despite technological advancements that have increased average yields. Greater precipitation partially mitigates damages from extreme high temperatures, but the study finds no significant correlation between temperature and precipitation outcomes in the raw daily data. The estimated climate-change impacts are insensitive to the specified growing season and consistent with time separability. The study combines fine-scale weather data with flexible regression models, finding consistent nonlinear temperature effects across time, locations, crops, and variations in temperature and precipitation. The study also notes that the estimated relationship is consistent when comparing estimates based on year-to-year weather variations and cross-sectional climate variations. This suggests limited historical adaptation to extreme heat for any given crop. The findings highlight the importance of considering nonlinear temperature effects in climate change models and the need for further research on potential adaptations to climate change, such as changing crop locations or seasons. The study also acknowledges the limitations of its approach, including the inability to account for CO₂ concentrations, which may affect plant growth and yields.