Abstract: Taking 20 overwintering Chinese cabbage seedlings as experimental materials, the grey correlation analysis method was used to analyze the morphological characteristics of Chinese cabbage seedlings and select suitable morphological characteristic indicators, in order to establish a prediction model for the overwintering morphological characteristics of Chinese cabbage seedlings in seed production.The results showed that the morphological characteristics of Chinese cabbage seedlings before winter, such as above-ground fresh weight, leaf area per plant, leaf number, rhizome thickness, and lateral root number, were highly correlated with the overwintering rate with the correlation coefficients of 0.798 1,0.750 6,0.735 2,0.707 4,and 0.676 9,respectively),and all indicators fully reflected the overwintering ability of the seedlings.Using the relative value of the selected ‘pre winter’ morphological indicator as the independent variable, we constructed a prediction model Y=68.885+0.963X₁-0.011X₂-0.168X₃+10.983X₅+0.116X₇.Through back testing, the accuracy rate of the model was more than 97.00%.Then we used the root mean square error（1.405 3） and relative error（0.014 8） to evaluate the model, and its predicted values were highly accurate, which could effectively predict the overwintering rate of Chinese cabbage seedlings in seed production.Overall, establishing the model using the grey correlation analysis method was reliable enough and could be used to predict the overwintering rate of Chinese cabbage seedlings in the open field, which provided strong reference for agricultural sci-tech staff.