The aim of this research was to conduct an empirical assessment of the risks of N–loss to groundwater associated with land use (LU), based on annual data on the net N–balance surplus in Spanish crops. These data were used to generate a detailed risk rating system reflecting the potential risks of N–loss from agriculture. The new LU ratings were used to assess the specific vulnerability of groundwater to nitrate pollution, by using the LU–IV procedure (Arauzo 2017).The study area included the catchment areas of 12 alluvial aquifers associated to tributaries of the Ebro River (Spain). Most of the alluvial aquifers were chronically polluted by nitrate, with only a few remaining unaffected by pollution. The LU maps from two different base maps (MCAE 2000–09; SIOSE 2011) were used to generate the respective versions of the map of vulnerability to nitrate pollution using the LU–IV procedure. Potential nitrate vulnerable zones (NVZ) were extracted from different models of vulnerability for comparison with the map of groundwater nitrate content. The models compared were the following: model A (LU–IV procedure, based on MCAE 2000–09 and using LU ratings from N–surpluses in Spanish crops), model B (LU–IV procedure, based on SIOSE 2011 and using LU ratings from N–surpluses in Spanish crops), model C (LU–IV procedure, based on MCAE 2000–09 and using LU ratings from bibliographical references; Arauzo, 2017), model D (IV index), model E (DRASTIC index), and model F (GOD index).Results confirmed, as expected, that models A and B proved to be the best risk predictors, both for polluted groundwater areas and for areas at risk of being polluted. These results support the high level of reliability of the LU–IV procedure, when applying the LU ratings obtained empirically from the N–surpluses.