Regional environmental risk assessment can be defined as risk assessment which deals with a spatial scale that contains multiple habitats with multiple sources of many stressors affecting multiple endpoints. The characteristics of the landscape also affect the estimated risk. In this paper, an information diffusion method based on a grid system is proposed to assess regional environmental risk. The risk information on a single environmental risk source can be diffused effectively by using fuzzy set theory. Regional environmental risk values obtained from information diffusion can be clustered on classification criteria and different environmental risk levels can be depicted in a spatial partition map. Huangge Town and Nansha Town located in the Nansha Area of Guangzhou City in China were selected as model cases. The results derived from this information diffusion method will help the local government to optimize the distribution of industrial areas and establish risk prevention measurements and emergency management procedures.