The investigation of treatment effect and safety profiles in pediatric or small population(s) have become more and more common nowadays in drug development, for reasons such as regulatory requirement (e.g. PREA/PIP), precision medicine (biomarker-enriched population) or the target indication (rare disease), etc.. Depending on the knowledge of the investigative therapy, statistical methods may be available to help leverage information that are internal/external to a study, and thus can limit drug exposure in pediatric subjects, or overcome challenges caused by small population size. For example, one can use Bayesian approach to augment pediatric studies by the benefit/risk assessment obtained from the adult population, use Basket trial design to share information concurrently across biomarker-defined subpopulations, or use historical control to serve as the comparator when studying (ultra-)rare disease. In this roundtable, we will review some case studies from the above mentioned scenarios and motivate the discussions and experience sharing among attendee, where we learn up-to-date statistical methodologies and good practice from each other.