Abstract: To make up for the defect of sensory method for evaluating green tea taste quality, chemistry instrument and electronic tongue were used for the quantitative evaluation of green tea taste quality. Ten main taste compositions of tea polyphenols, amino acids, caffeine, GA, EGC, C, EGCG, GCG, ECG and total content of catechins were determined by chemistry instrument. The sensor signals of green tea infusion of each sample were obtained by electronic tongue. The BPANN models of 10 taste composition contents with sensory taste scores and sensor signals with sensory taste scores were established, respectively. The results show that the predicting results of green tea sensory taste quality by chemistry instrument and electronic tongue are reliable. However, the performance of BPANN model with RMSEP of 1.913 and Rp of 0.932 in the prediction set by electronic tongue is better than that by chemistry instrument. The results demonstrate that electronic tongue can better predict the sensory taste quality of green tea.