The premise of the implementation of eligible pharmacodynamic biomarkers (PD markers) in clinical development of drugs is based on their qualification and understanding of human disease networks on a molecular level, which may be relevant to risk factors, pathogenesis, prognosis, and relapse/remission. Especially, information on PD markers characterized with drug exposure in target tissues, drug binding to target molecules, and linkage to clinical endpoints in early drug development stage can be critical for GO/NO GO decision for the next late clinical stages. Moreover, early confirmation of reliable biomarker method validations consisting of analytical performance and sample handling performance classified with fit-for-purpose strategy would be more crucial in practice for trouble-free biomarker implementation. For clinical setting of PD markers and final success of drug regulatory approval, good interpersonal communications among various members such as medical-, biological-, pharmacological-, toxicological-, pharmacokinetics-, statistical-scientists, and bioanalysts are also required. We are now trying to establish a knowledge-based biomarker selection method using commercially available databases, and our policy of fit-for-purpose-based biomarker method validation. In this article, we will report our current thinking and case-studies mentioned above.