Curing cancer requires the treatment of metastatic disease. Whether this is a patient with advanced disease and clinically apparent metastases, or if the patient with localized disease is at risk for development of dissemination, failure to control metastasis will result in a poor outcome. Here, we have presented a molecular guide to our current understanding of the processes underlying metastasis. Experimental clinical trials designed to further the understanding of metastasis are often limited by selection of patients with advanced disease. Therefore, our understanding of the processes involved in the metastatic cascade is limited by the availability of comprehensive experimental model systems. The study of metastasis relies most heavily on xenografts, tumors using human cell lines, or tumor tissue that can grow in mice. These models present a limited recapitulation of the patients. Xenograft models require some degree of immunosuppression on the part of the host, because mice with native immune systems will reject transplanted human tumors, preventing their growth. As a result, mice with immune defects ranging from depleted T cells (nude mice) to absent T, B, and NK cells (SCID-Beige) are used as hosts. As the evasion of the immune system is a key function demonstrated by the metastatic cancer cell, xenograft models, by necessity, subvert this step. Furthermore, recent studies have established that angiogenesis in transplanted tumors is different than in native tumors, further highlighting the limitations of these models. With these limitations, studies of metastasis may require development of models of autochthonous tumors, that is, tumors originating in the study animals. A number of cell lines of autochthonous murine tumors have been established that generate metastatic disease after implantation into mice. Moreover, some transgenic animals spontaneously develop metastatic tumors that, although occurring in genetically engineered animals, may represent the most complete model from early development to late effects. Finally, a very promising field of autochthonous tumor studies lies in work with companion animals (pets). Some dogs will have cancer, often with striking similarities to those of their human counterparts. These pets may represent an important study group, because they have autochthonous tumors, occurring spontaneously, in an outbred population. In all of these cases, the tumor, new vasculature, and the immune system are syngeneic with the host. In addition to the advances in model systems, advances in technology will further our understanding and ability to combat metastatic disease. As demonstrated, genomics is proving to be a powerful tool in identifying those at risk for metastasis. From these genetic signatures, molecular targets may be deduced from the genes altered in patients with poor prognoses. Furthermore, other molecular tools such as proteomic analysis may provide further information. Clearly, therefore, a synthesis of different technologies and complimentary information will be required to target metastases and improve the outcome for patients affected by them.