This research used multivariate statistical analysis to evaluate the tolerance of 11 tomato cultivars to low light at the seedling stage. The low-light condition (200–420 μmol/m2 s) was simulated by a shade net. It was found that 11 of 16 character indices of different cultivars, such as dropping angle, bend degree, and accumulation of leaf area, showed a significant difference by using multiple variance analysis. After factor analysis, the 11 character indices could be summarized into 5 main factors with a cumulative contribution rate of 84.968 %. According to the factor scores after varimax rotation, the 11 tomato cultivars could be classified into three categories by using cluster analysis. The severely low-light-sensitive cultivars were T1, T5, T6, T10, and T11 and the moderately low-light-sensitive cultivars were T4, T7, and T9. Cultivars T2, T3, and T8 were resistant to low light. In accordance with the appraisal result, the light-sensitive cultivars T5 and T10, the moderately low-light-sensitive cultivar T4, and the low-light-tolerant cultivar T8 were randomly selected to observe the variation in the ultrastructure of leaves of different tomato cultivars with the aid of a transmission electron microscope (TEM). In chloroplasts of T5 and T10, membranes were heavily damaged and mitochondria were vacuolated, whereas the chloroplast structure of T4 was slightly damaged and its mitochondria grew normally. In the chloroplasts of T8, the organelle membranes were intact, the degree of thylakoid stacking was high, and mitochondria grew normally. Our results showed that multivariate statistical analysis of low-light tolerance in tomatoes has certain scientific applicability.