Abstract: In this paper, for resolving the flaws showed in the conventionally used sensory evaluation method, the authors presented a new approach to solving the problem of selecting proper and excellent cider yeast. The eight different yeast strains commonly used in the cider industry were selected for this experiment. The sensory evaluations of different ciders, which were fermented by the yeasts, were done first , then their aroma compounds were identified by the Headspace-Solid phase microextraction-Gas Chromatography-Mass Spectrometry, and their contents were determined using quantitative standards with known amounts of target compounds and internal standards.A model of Fuzzy Comprehensive Evaluation was established and was used to assess and rank the cider-making attributes of all eight yeast strains. F14 was selected out of eight yeasts for the best cider by Fuzzy Comprehensive Evaluation Method. The results indicate that the method in this paper uses more cider sample information and it is more accurate and more objective than conventional methods.