Abstract: Abstract: In this study, we aimed to establish an optimal index model of tomato seedlings to improve the accuracy of seedlings evaluation. Based on the determination of eleven individual indicators such as plant height, stem circumference, root-shoot ratio (Fresh), plant fresh weight, root-shoot ratio (Dry), plant dry weight, chlorophyll a, chlorophyll b, total chlorophyll, carotenoid and SPAD of about 300 tomato seedlings with three leaves, according to the seedling standard, the membership function formula of each index was established and the single factor evaluation matrix was calculated. The weight coefficient matrix was calculated according to multiple linear regression. Finally, the fuzzy comprehensive evaluation method was used to multiply the single factor evaluation matrix and the weight coefficient matrix to calculate the comprehensive evaluation matrix and establish a comprehensive evaluation index of tomato seedlings. This index can stably and accurately evaluate tomato seedlings, but the calculation was relatively complicated. Therefore, a simple and effective seedling model was needed instead of the comprehensive evaluation index. Then the principal component analysis was used to screen the indicators and combine them into index model of tomato seedlings. After that, the suitable model was selected through the correlation analysis between the comprehensive evaluation index and seedling index model. Finally, the optimal model was selected based on the verification of suitable model using industrialized nursery seedlings of three different tomato varieties. The results showed that the weight coefficients of each indicator such as plant height, stem circumference, root-shoot ratio (Fresh), plant fresh weight, root-shoot ratio (Dry), plant dry weight, chlorophyll a, chlorophyll b, total chlorophyll, carotenoid and SPAD of tomato seedlings were 0.180, 0.085, 0.098, 0.076, 0.103, 0.078, 0.070, 0.070, 0.070, 0.072, and 0.097, respectively. And the index was divided into three subordinate functions: parabola, positive S and straight line. The parabola included plant height, the positive S included stem circumference, plant fresh weight and plant dry weight, and the straight line included root-shoot ratio (Fresh), root-shoot ratio (Dry), chlorophyll a, chlorophyll b, total chlorophyll, carotenoid and SPAD. The comprehensive evaluation index of tomato seedlings was calculated to be 0.26-0.79, which could comprehensively summarize the overall quality of seedlings as well as objectively and accurately evaluate the robustness degree of seedlings. The eleven individual indicators of seedlings could be divided into three principal components, including morphological indicators, pigment indicators and plant height. The morphological indicators included stem circumference, root-shoot ratio (Fresh), plant fresh weight, root-shoot ratio (Dry) and plant dry weight. The pigment indicators included chlorophyll a, chlorophyll b, total chlorophyll, carotenoid and SPAD. Plant height was divided into one component. Twenty-five seedlings index models were combined using one indicator from each principal component and taking (stem circumference/plant height) × whole plant dry weight and (stem circumference/plant height + root dry weight/dry weight of the ground) × whole plant dry weight as the control group, furthermore, four suitable models such as (chlorophyll a/plant height) × whole plant dry weight, (total chlorophyll / plant height) × whole plant dry weight, (carotenoid/plant height) × whole plant dry weight and (SPAD/plant height) × whole plant dry weight were selected due to the high correlation and representativeness. According to the model verification results of three different tomato varieties, (chlorophyll a / plant height) × whole plant dry weight and (total chlorophyll/plant height) × whole plant dry weight were identified as the optimal models to evaluate the quality of tomato seedlings because of high correlation with the comprehensive evaluation index, good stability and representativeness. This result showed that the fuzzy comprehensive evaluation was good for assessment of tomato seedlings. According to the comprehensive evaluation index and cluster analysis of the selected seedling index model of tomato seedlings, the quality of tomato seedlings was divided into three grades, and the range of seedling index of different grades was given. The research can provide theoretical basis for tomato seedling production and also provide an optimized reference for evaluating quality of other vegetable seedlings.