The seminal 1993 article by LaFollette and Shanks "Animal Models in Biomedical Research: Some Epistemological Worries" introduced an influential taxonomy into the debate about the value of animal experimentation. The distinction they made between hypothetical and causal analog models served to highlight a concern regarding extrapolating results obtained in animal models to human subjects, which endures today. Although their taxonomy has made a significant contribution to the field, we maintain that it is flawed, and instead, we offer a new practice-oriented taxonomy of animal models as a means to allow philosophers, modelers, and other interested parties to discuss the epistemic merits and shortcomings, purpose, and predictive capacities of specific modeling practices.