Based on the spectral characters of corn leaf nitrogen content in the space, the spectral models for rapid estimating crop nitrogen content were set up, which is practically meaningful to effectively providing the guidance in fertilization. Spectral technology was applied to explore corn leaves nitrogen content distribution regularity and the relationship between the nitrogen content and plant index was analysed and then the estimation models were built. The results showed N content in upper leaves is higher than that in lower leaves in four growing stages; lower leaves at tassel emerge stage are sensitive to nitrogen losses, which could be used in guiding fertilization in grain production; optimum estimation models were built atjointing stage, the full-grown stage and tasseling stage, The research results provided the proof of crop nutrient analysis and rational fertilization.