Since the earliest introduction of cytotoxic chemotherapy, investigators have pursued laboratory techniques designed to match patients to available drugs. Most of the work, published through the 1980s, reflected the prevailing view of cancer as a disease of dysregulated cell proliferation. Noteworthy, the description of apoptosis and programmed cell death, fundamental to our modern understanding of human tumor biology, did not occur until well after the heyday of in vitro chemosensitivity testing. By incorporating the modern tenets of carcinogenesis associated with perturbations in cell survival we can now re-examine laboratory assays of drug response in the context of drug-induced programmed cell death. Although there is interest in the use of genomic analyses for the prediction of chemotherapy response, the painful recognition that genotype does not equal phenotype will continue to limit broad application of these platforms. Biosystematics instructs that biological pathways rarely follow predicted routes. Efforts to force human biology to behave according to preconceived scientific dictates have proven costly and unsuccessful. Whole-cell experimental models with the capacity to evaluate all the operative mechanisms of cellular response to injury, acting in concert, provide valid tools for the study of human cancer. Educated by cellular behavior, we can expeditiously examine molecular processes of interest. This article briefly reviews the history of whole-cell experimental models of in vitro chemosensitivity testing then focuses on cell-death measures as the most robust predictors of clinical outcome in human cancer.