The main purpose of this article is to investigate the persistence of fungicides in the environment. QSAR models using four types of reactivity descriptors were constructed to predict the degradation rate constants and examine chemical interactions, to further assess and classify the environmental risks of fungicides. Two major findings emerged. First, the model results show that the degradation in surface water of fungicides is mainly affected by the polarization. The maximum nucleophilic condensed local softness is the most important descriptor. Second, both polarization and chemical potential affect degradation in the soil. The maximum electrophilic condensed local softness is the most important descriptor. The findings not only identified 20 kinds of high ecological risk fungicides, but also showed that phthalimides, sulfamides, and antibiotics are less harmful to the environment because of low persistence and low bioconcentration factors. This approach provides a basis for interpreting chemical interactions between fungicides and environment.