This paper studies a critical margin of agricultural adaptation to climate change: how farmers’ planting decisions have adapted to climate change in the US. By linking a newly-compiled panel of profits and yields for nine major crops with county-level temperature data, I provide the empirical estimate of the cropland allocation elasticity to climate change. I design a novel two-stage method that separates the temperature-affected yields from profit and other unobserved factors under a discrete choice model. The estimation yields an allocation elasticity of 2.5, suggesting that a 1% increase in the temperature-affected yield ratio would lead to about a 2.5% increase in the land allocation ratios between two crops. When elasticities are estimated by individual annual cross-sections, the allocation elasticities display significant correlations with the national growing-season average temperature, suggesting decreasing adaptation intensity with rising temperatures. Our findings reveal a trade-off where farmers are allocating land to crops with less damage from the weather rather than higher yields, indicating passive adaptation. Cropland allocation alone is insufficient to mitigate the damage caused by climate change, highlighting the importance of other adaptation and mitigation practices.