Abstract：According to classification criterion of highway tunnel，the classification indexes are comprehensively considered，and neural networks method of rock mass classification of highway tunnel is established. Proper classification indexes are chosen and standardization of indexes data are conducted by vector method，and training samples of neural networks are built. Training results of neural networks can provide the judgement basis for rock mass classification of highway tunnel. Hereby，surrounding rock mass in Hanjialing tunnel is classified. The feasibility，veracity，and advantage of founded neural networks method are verified.
