Toxicity studies in mammals continue to be the most appropriate model for predicting risk in humans, but they tend to be expensive and time-consuming. In the aftermath of the genetic sequencing of zebrafish (Danio rerio), that species showed to be highly genetically homologous to humans. The use of the zebrafish model to assess food toxicity is already a reality as it is capable of biological processes difficult to reproduce in vitro. Studies of complex mechanisms of absorption, distribution, metabolism and excretion as well as cellular and tissue interactions are of great information value resulting in time, space and cost savings, when compared to studies with rodents. This review addresses the relevance of zebrafish model in food safety research, both in the use of ingredients and innocuous food additives as well as for establishing levels of safe food contaminant residues present in the environment. Toxicological screening using the zebrafish model integrate the evaluation of teratogenicity, cardiotoxicity, hepatotoxicity, genotoxicity, neurotoxicity, endocrinetoxicity, reproductive and behavioral aspects. These are important endpoints for food safety assessment, which take substantially less time than in mammalian tests. Furthermore, it serves well as a screening test follow-up for validating favorable results in murine models, hence accelerating the risk assessment process of products submitted for approval and registration, prioritizing safe compounds and reducing unnecessary costs in subsequent mammalian studies. In conclusion, zebrafish model can be a useful tool for food safety tests, however, additional studies are needed to further validate this model for registration of new food ingredients and additives.