To clarify the influence of different stewed times on the sensory quality of mutton soup， 5 mutton soup samples prepared with different stewed times（DTS） were evaluated by sensory fuzzy integrate evaluating method （SFE） and value engineering evaluation（VEE）， and the volatile compounds were identified by gas chromatography-massspectrometry（GC-MS）. On this basis， partial least squares regression（PLSR） was established among volatile components， SFE and VEE scores to explore the contribution of volatile components to the sensory properties of DTSs in an intuitive form. Results showed that 58 volatiles were identified in DTS samples， of which could be classified into 4 value functional groups according to their flavor contribution， namely Z1 （terpenes and alcohols）， and Z2 （aldehydes）， Z3 （phenolic compounds and acids）， and Z4（ketones， esters and ethers）. Function index ranked in the order： DTS4 > DTS5 > DTS3 > DTS2 > DTS1， which was in accordance with the SEF results. Thus， 4 h was the best stewed time for preparing mutton soup. At the same time， PLSR indicated that functional components， such as 3-methyl-butyraldehyde， （E）-2-decenaldehyde， and carvacrol were significantly positively correlated with the SFE score， showing a positive effect on DTS flavor； while 4-methyl phenol and 4-methyloctanoic acid showed significantly negative correlation with SFE， thus reducing their content might improve SFE score of mutton soup.