Differently from the conventional antineoplastic agents, target-based drugs are designed a priori, based on our knowledge of various physiological molecules that has been obtained by the development of molecular biology. This "Copernican revolution" in drug development may imply a paradigm shift in this field. However, contrary to the initial expectations, many drugs developed by this approach are now faced with difficulties, mainly because of the fundamental and theoretical limits of this approach. All of the physiological functions are not always known in all target molecules. In low-molecular-weight drugs, i.e., "inhibitors," targets disperse, due to the structural similarities in physiological molecules. This double-faced "out-of-focusing" causes many problems in various steps of drug development, drug design, clinical trials, and administration to patients. Many drugs are now being abandoned because of unexpectedly lower response rates or unforeseeable adverse effects, and the variety of the drugs exhibits a kaleidoscopic appearance. The double-faced "out-of-focusing" derives from the methodological limits in molecular biology, i.e., elementalism, and limits in our techniques for drug development. To overcome these currently inevitable limits, it appears essential to elucidate the specific changes in target molecules that chiefly promote tumor growth and, consequently, strongly predict response to the administered drugs. Precise and efficient detection of responder populations is the key to the development and establishment of target-based anticancer therapies.