“Hongyan” strawberry was selected as test materials, with the storage temperature of 4 ℃, humidity of 85%~95%, to detect the strawberry freshness based on flavor. Electronic nose (e-nose) and the single ethanol sensor were used to collect the odor of strawberry during storage period. The strawberry was divided into four stages based on the changes of sensory evaluation score and Principal component analysis. By using the partial least squares discriminant analysis (PLS-DA) and classification support vector machine(SVM-C), the model of fresh degree of strawberry was established based on odor discrimination. The overall accuracy for modeling group and validation group were respective 84.2% and 88.3% based on PLS-DA. The overall accuracy for modeling group and validation group were 99.2% and 97.7% respectively, based on SVM-C method. The overall accuracy for modeling group and validation group were 83.3% and 86.70% based on PLS-DA method, 90.8% and 90.0% based on PLS-DA method. This research can supply technical help to monitor the freshness of strawberry change during storage and transportation.