Cancer can largely be conceived as a consequence of genomic catastrophes resulting in genetic events that usurp physiologic function of a normal cell. These genetic events mediate their pathologic effects by either activating oncogenes or inactivating tumor-suppressor genes. The targeted repair or inactivation of these damaged gene products may counteract the effects of these genetic events, reversing tumorigenesis and thereby serve as an effective therapy for cancer. However, because they are the result of many genetic events, the inactivation of no single mutant gene product may be sufficient to reverse cancer. Despite this caveat, compelling recent evidence suggests that there are circumstances when even the brief interruption of activation of a single oncogene can be sufficient to reverse tumorigenesis. Understanding how and when oncogene inactivation reverses cancer will be important in both defining the molecular pathogenesis of cancer as well as developing new molecularly based treatments.