Widespread reliance on animal models during preclinical research and toxicity testing assumes their reasonable predictivity for human outcomes. However, of 20 published systematic reviews examining human clinical utility, located during a comprehensive literature search, animal models demonstrated significant potential to contribute toward the development of clinical interventions in only two cases, one of which was contentious. Included were experiments expected by ethics committees to lead to medical advances, highlycited experiments published in major journals, and chimpanzee experiments—the species most generally predictive of human outcomes. Seven additional reviews failed to demonstrate utility in reliably predicting human toxicological outcomes such as carcinogenicity and teratogenicity. Results in animal models were frequently equivocal, or inconsistent with human outcomes. Consequently, animal data may not generally be considered useful for these purposes. Regulatory acceptance of non-animal models is normally conditional on formal scientific validation. In contrast, animal models are simply assumed to be predictive of human outcomes. These results demonstrate the invalidity of such assumptions. The poor human clinical and toxicological utility of animal models, combined with their generally substantial animal welfare and economic costs, necessitate considerably greater rigor within animal studies, and justify a ban on the use of animal models lacking scientific data clearly establishing their human predictivity or utility.