Oncogenesis occurs through the acquisition and selection of multiple somatic mutations--each contributing to the growth, survival and spread of the cancer. Key attributes of the malignant phenotype, such as unchecked proliferation and cell survival, can often be "reversed" by the selective diminution of dominant oncogenes by chemical or genetic means (e.g. beta-catenin in colorectal carcinomas; bcr-abl in chronic myelogenous leukemias (CMLs)). These observations suggest that the products of oncogenes, or of secondary genes that mediate and maintain tumor phenotypes, might be revealed through the systematic disruption of each and every gene in tumor-derived cells. Some of these genes may encode proteins amenable to therapeutic intervention, thus fueling the cancer drug discovery process. However, a functional assessment of each known or predicted gene in mammalian cells is a daunting task and represents the rate-limiting step in drug target identification and validation. In this regard, RNA interference (RNAi) by small interfering RNAs (siRNA) holds great promise as the "tool of choice" to mediate the selective attenuation of mammalian gene expression and protein function. Here, we review strategies by which RNAi might be used to determine the genetic alterations that contribute to malignant transformation via large-scale cell-based screens, and propose how this information can be used in conjunction with small molecule screens to identify pathways critical to cancer cell survival.