At present, significant resources are committed by both academic research and industry to deliver groundbreaking therapeutic and diagnostic strategies aimed at curbing cancer occurrence. Despite the critical mass of available knowledge and technology platforms in the anti-cancer drug development field providing for the development of pathway-targeted therapies, only a few efficacious treatments have been developed. The major preclinical point of the government-regulated process for extensive animal testing of potential anti-cancer therapeutic compounds is directed towards safety assessment prior to the clinical introduction of any new anti-cancer drug. This analytical path of demonstrating the desired drug efficacy while proving it to be non-toxic has been implemented in a number of countries as multiphase clinical trial procedures. The initial stage in drug candidate evaluation demands routine investigation of future therapeutic compounds in animal and cell culture models, primarily to obtain knowledge of non-toxicity ranges, interaction with metabolic pathways, and systemic pharmacological behavior. This phase, termed preclinical characterization, may also provide a fair estimate point for the compounds therapeutic efficacy given the availability of appropriate disease model(s). In general, considering the extremely long (routinely >10 years) and expensive (often in excess of $700 million per drug lead) process to obtain the regulatory approval to market therapeutic compounds, the quality and the scope of efficacy data obtained during the preclinical stages may expedite clinical testing, dramatically increase affordability of downstream drug development steps, and advance the identification of effective cancer therapies. Currently, drug efficacy studies are conducted almost exclusively in xenograft models that employ transformed human cell lines to initiate tumor growth upon injection into immunocompromised animals. Though easily derived, the xenograft models feature multiple intrinsic limitations that jeopardize the predictability of drug testing output data. Xenograft tumors are developed from a genetically heterogeneous cell population that has been maintained in vitro for multiple passages. Moreover, the tumor growth occurs in an ectopic, non-physiological environment in the absence of immune surveillance and systemic interactions with the vascular system. As an alternative source of experimental tumors, animal models of spontaneous carcinogenesis may be also employed, but these models generally lack the reproducibility in timing of tumor onset and feature by heterogeneous tumor characteristics/drug response due to a considerable genetic "noise" caused by non-inbred strain background. The drawbacks of the xenograft and the spontaneous tumorigenesis models are largely ameliorated in genetically engineered mouse (GEM) tumor lines, which provide preclinical researchers with the ability to study naturally occurring tumors featuring pathway aberrations typical for similar human cancer types in the context of an appropriate tissue environment in immunocompetent animals. This approach finds ground in the rapidly expanding knowledge base for molecular mechanisms underlying carcinogenesis in human patients and is further fueled by the recent and rapid progress in the methodology and availability of resources to design and construct sophisticated animal models for a broad spectrum of human malignancies. At present, the GEM strategy not only provides an opportunity to combine in transgenic animals multiple genetic aberrations closely matching those detected in human patients, but also the potential to interfere with cancer-related molecular pathways in both a tissue-restricted and time-specific manner, providing genetic evidence for drug target legitimacy. This translates into a more accurate prediction of the dynamics of tumor progression while minimizing individual variations. In contrast to xenograft models, the cancerous lesions that occur in GEM animals exhibit a high degree of genetic similarity due to the availability of inbred and congenic lines for GEM generation. Among other benefits, this genetic similarity allows the decoding of individualized "tumor molecular signatures" that may be applicable for identifying cancer prognostic markers and developing personalized anti-tumor therapies based on predicting an individuals response to drug treatment known as the patient stratification principle. The CCR Center for Applied Preclinical Research (CAPR), will develop and implement a comprehensive preclinical trial framework for evaluating the anti-tumor efficacy and selectivity, safety, biodistribution, and metabolism of early stage candidate drugs using GEM models. CAPR will establish the infrastructure required for preclinical evaluation of anti-cancer drug leads in a recently derived collection of GEM models for high-occurrence cancers such as lung, ovary, and prostate gland tumors, as well as more rare cancer types that fall under the "unmet" demand category such as high-grade astrocytomas, hepatocellular carcinomas, and pancreatic cancers. CAPR will also be responsible for continuous sampling of the dynamic knowledge base on the mechanisms of carcinogenesis to seek additional molecular targets related to the tumor formation process. The appropriate GEM strains will be designed, derived, or adapted to address the CAPRs internal demand for novel mouse tumor models;however, these resources will also be shared with other scientific and industrial communities engaged in anti-cancer drug discovery and development.