This article provides an overview of the animal models currently used in tuberculosis research, both for understanding the basic science of the disease process and also for practical issues such as testing new vaccine candidates and evaluating the activity of potential new drugs. Animals range in size, from zebrafish to cattle, and in degrees of similarity to the human disease from both an immunological and pathologic perspective. These models have provided a great wealth of information (impossible to obtain simply from observing infected humans), but we emphasize here that one must use care in interpreting or applying this information, and indeed the true art of animal modeling is in deciding what is pertinent information and what might not be. These ideas are discussed in the context of current approaches in vaccine and drug development, including a discussion of certain limitations the field is currently facing in such studies.