Abstract:With the development of spectral technology and its inherent advantage,more and more researches were carried out to investigate the application of spectral technology on plant nutrition diagnosis and nutrients contents prediction.In the article,firstly,the principle and physiological basis for plant spectral diagnosis was briefly introduced.Then the methods of hyperspectral data extraction and analysis were summarized focusing on the crops nitrogen,phosphorus,potassium and other nutrients nutrition diagnosis and nutrients contents prediction.A big research progress has made in the research concerning N nutritional spectral diagnosis,in which sensitive wave bands were put forward and spectral indices were constructed and validated.However,the research concerning P and K nutritional spectral diagnosis was much less comprehensive and the result were inconsistent;other nutrients spectral diagnosis were touched upon in some sort and much research should be continued further.In addition,a detail analysis for the existing difficulties about spectrum data collection and nutrients prediction models was performed.Finally,the future advance trends and the prospect of application were brought up.