Environmental risk assessment is an essential element in any decision making process in order to minimize the effects of human activities on the environment. Unfortunately, in many occasions, environmental data tend to be vague and imprecise and as a consequence uncertainty is associated to any study related to it. Uncertainty in risk assessment may have essentially two origins: randomness and incompleteness. In this paper, fuzzy logic is used to manage this uncertainty in environmental data concerning accidental releases in chemical plants. The methodology developed allows assessing the environmental risk of such releases using a set of parameters treated with fuzzy logic. This method can be used as a powerful tool by both public authorities and plant managers to take decisions in situations where chemical releases can occur.