Abstract: A new method about detecting the caffeine and five kinds of catechin components including Catechin,Epicatechin,Epigallocatechin,Epicatechin gallate,Epigallocatechin gallate of Congou black tea by near infrared spectroscopy was established.There were 240 test samples,180 samples of them used to be a correction stage as the remaining 60 samples a prediction set.Each model was optimized the best waves of the number of segments and best pretreatment method for model inguse to establish the quantitative prediction model by OPUS 7.0software.The Smooth points were 17 and dimension was 1. Combined with caffeine and five kinds of catechin component induding C,EC,ECG,EGC,EGCG content,the prediction model was established,and the performance of the prediction model was analyzed.The model predicted a high accuracy which can be used to predict the caffeine and five kinds of catechin component including C,EC,ECG,EGC,EGCG quality of Congou black tea.The calibration correlation coefficient( Rc) was 91.85% ~99.49%,correcting root mean square error( RMSEC) was 0.0187~ 0.353; predictive correlation coefficient( Rp) was 97.12% ~ 99.88%,and the RMSEP was 0.00759 ~ 0.0773. Each model calibration set and prediction set had a higher degree of fit,the prediction performance model of caffeine >EGC > EC > ECG > EGCG > C.The combination of near- infrared spectra of each prediction model to predict the performance had an excellent organoleptic results and can be used in the caffeine and five kinds of catechin component C,EC,ECG,EGC,EGCG content rapid non- destructive testing of congou black tea.