At present, the starting evaluation method of Dual Clutch Transmission (DCT) vehicles is mainly based on the driver’s subjective evaluation. Too much subjectivity leads to inaccurate evaluation. In view of the above problems, a DCT vehicle starting quality evaluation system was established based on the grey relational analysis method and the fuzzy comprehensive evaluation method. In this study, power, comfort and economy were considered, and starting quality evaluation indicators including starting response delay time, starting acceleration time, starting acceleration peak, engine speed fluctuation and shock were selected as evaluation index. The coefficient of variation method calculated the objective weight of the starting quality evaluation indicators, and based on the grey relational analysis method, the order of the correlation degree between each index and the starting quality was clarified, which verified the accuracy and reliability of the weight calculation results. Finally, based on the fuzzy comprehensive evaluation, the starting quality evaluation model of DCT vehicles was established and objectively scored, which made the starting quality evaluation of DCT vehicles more objective and true.