Animal models have been instrumental in elucidating key biochemical and physiologic processes of cancer onset and propagation in a living organism. Most importantly, they have served as a surrogate for patients in the evaluation of novel diagnostic and therapeutic anticancer drugs, including radiopharmaceuticals. Experimental tumors raised in rodents constitute the major preclinical tool of new-agent screening before clinical testing. Such models for oncologic applications today include solid tumors raised in syngeneic fully immunocompetent hosts and human xenografts induced in immunodeficient mouse strains, and tumors spontaneously growing in genetically engineered mice represent the newest front-line experimental modality. The power of these models to predict clinical efficacy is a matter of dispute, as each model presents inherent strengths and weaknesses in faithfully mirroring the extremely complex process of human carcinogenesis. Differences in size and physiology, as well as variations in the homology of targets between mice and humans, may lead to translational limitations. Other factors affecting the predictive power of preclinical models may be animal handling during experimentation and suboptimal compilation and interpretation of preclinical data. However, animal models will remain a unique source of in vivo information and the irreplaceable link between in vitro studies and our patients.