Remote sensing offers the possibility to obtain a rapid and non-destructive diagnosis of crop health status. This gives the opportunity to apply variable rates of fertilizers to meet the actual crop needs at every locations of the field. However, the commonly used normalized difference vegetation index (NDVI) based on canopy reflectance shows delayed response to nitrogen deficiency (i.e. more reliable from V8 growth stage of corn) and saturates at canopy closure. Our objectives are to determine if by modifying the sensing geometry, NDVI could provide earlier N-status of crop, to determine effect of averaging the readings over a larger scale could reduce the noise and increase correlation with crop N-status and to determine if satellite based NDVI saturates later than proximal based NDVI. Results will be shown at the conference.