More than 90% of new drugs have failed in clinical trials. We first examined the characteristics of failed new drugs for the treatment of seven types of cancers. We then reviewed the hazard ratios of survival in typical phase III clinical trials of drugs for these cancers from the last five years. The data suggested that population sizes in most clinical trials were limited to a certain level of detection. Evidently, each drug has its effect only in certain individuals with defined characteristics, and the success and failure of a new drug depends on these characters such as ethnic group, sex, environmental condition, pathological feature, and genotype. Due to the complexity of influence of multiple factors and the current limitation in understanding of them, tremendous number of subgroups among patients may have not been recognized. Therefore, decision based on only a few of statistical comparisons may not always provide a correct judgement for a new drug. The drugs that do survive clinical trials are winners regardless of how differences in genotypes or other characteristics effected patients in comparisons between new drugs and placebos, or between new and existing drugs. While drugs that are effective on certain characterizations or specific group of patients often are categorized as failure in the clinical trials based on the current statistical criteria. Thus, previously failed drugs can be reevaluated and reutilized by analyzing whether these drugs have different effects on various genomic populations, or on patients with other differing and variable characteristics.