This study explores whether the soil information contributes additional explanatory power beyond the base premium rate to crop insurance loss. By examining a panel of 697 counties from the Corn Belt states in the U.S. over the period of 2005—2015, we find the loss costs are systematically associated with the National Commodity Crop Productivity Index (NCCPI) conditioning on the county base premium rates, weather conditions, and year fixed effects. The loss cost rises first with NCCPI in lower NCCPI quartiles and then decreases in higher NCCPI quartile. In general, counties with medium NCCPI values are expected to have higher relative loss cost comparing with low and high NCCPI counties. The pattern of NCCPI’s effects is robust to different model specifications, heteroskedasticity and spatial correlation of data, as well as subsamples by insured acreage thresholds. The finding of this study presents empirical evidence that there is additional information embodied in soil and spatial variables not captured by the current base premium rates that is correlated with loss experiences. It suggests the potential to incorporate soil and spatial information to improve the crop insurance ratemaking.