The dominant paradigm in cancer treatment is that cancer cells die from the common pathway of apoptosis subsequent to DNA damage by anticancer agents and that cells resistant to apoptosis are resistant to therapy. In this review we trace the logic that brought about this view and discuss whether the clinical and experimental data that have now accumulated over the past decade support the position. We show that there is support for the apoptosis model only for certain malignancies of hematopoietic origin. For the majority of cancers (which are of epithelial origin), even though some may exhibit apoptosis after treatment, there is little or no support for the concept that apoptosis, and the genes that govern it, determine the response to therapy. In general, whether apoptosis matters for overall tumor response depends on how soon after treatment apoptosis occurs. If it occurs early (within 4 to 6 hours after treatment), it is likely to be important for determining the overall response of the cell or tumor. This is the case for some tumors of myeloid and lymphoid origin. On the other hand, if apoptosis occurs long after exposure, 24-48 hrs, and usually after mitosis, then it is unlikely to be the determinant of cytotoxicity, and modifying it is unlikely to affect overall cell killing. This is the situation for most, if not all, tumors of epithelial or mesenchymal origin. What then causes cancer cells to die following treatment with anticancer agents? The evidence from ionizing radiation is clear: DNA damage leading to chromosome breaks. We argue that the situation is similar for most anticancer drugs. Indeed inability to repair the DNA damage produced by these agents, and the levels of the proteins involved in these repair processes, may explain the treatment sensitivity of some cancers. Further clinical studies in this area are needed.