Crops in the coterminous United States are highly exposed to inland floods, yet large-scale estimates of crop-specific flood losses remain limited. Existing assessments are often geographically constrained and rely on a coarse empirical understanding of the relationship between flood duration and crop damage. To address these gaps, we developed a new nationwide library of frequency-duration-damage functions and applied it to determine the magnitude and spatial distribution of crop flood loss. The damage functions were empirically constructed by combining streamflow-based frequency-duration analysis with crop-specific duration–damage relationships from satellite-derived vegetation indices. We implemented the resulting functions in the HEC-FIA flood loss model, along with geospatial data on riverine and pluvial flood hazard, crop distribution, and crop production, to estimate economic losses for six major crops at watershed and state scales. Total annualized loss is approximately USD 258 million for a baseline June 1 flood scenario, with losses concentrated in the Upper Mississippi River Basin. Corn and soybeans account for over 85 % of total loss, while alfalfa, non-alfalfa hay, cotton, and rice experience smaller but regionally significant impacts. Relative loss patterns differ from absolute losses, with the highest proportional losses occurring in the Appalachian Valley, New England, and parts of the West. This study provides the first known national-scale assessment of crop-specific flood losses that integrates flood frequency-duration-damage relationships, enabling the identification of priority regions and crops for targeted adaptation. Future work should refine empirical frequency-duration-damage relationships and assess the sensitivity of crop loss estimates to model assumptions.