Introduction: The purpose of developing and adopting regulatory science (RS) for drug regulatory authorities (DRAs) is to enhance regulatory capacity by advancing the scientific approach for the evaluation of health-related products. While many DRAs around the world advocate the concept of RS, the implementation approaches of RS vary according to local needs and have not been systemically examined. This study aimed to systematically identify the evidence about how RS was developed, adopted, and advanced by the selected DRAs, and analyzed and compared the implementation experiences of RS development under the guidance of an implementation science framework. Methods: Documentary analysis of government documents and a scoping literature review were conducted, and data analysis was performed under the guidance of the PRECEDE-PROCEED Model (PPM). DRAs in the United States, the European Union, Japan, and China had officially launched RS initiatives and were therefore selected as the target countries in this study. Results: There is no common consensus on the definition of RS among the DRAs. However, these DRAs shared the same goal of developing and adopting RS, which was used to develop new tools, standards, and guidelines that could improve the effectiveness and efficiency of the risk and benefit assessment of the regulated products. Each DRA had decided its own priority areas for RS development and thus set specific objectives that might be technology-based (e.g., toxicology and clinical evaluation), process-based (e.g., partnership with healthcare systems and high-quality review/consultation services), or product-based (e.g., drug-device combination products and innovative emerging technologies). To advance RS, considerable resources had been allocated for staff training, advancing information technology and laboratory infrastructure, and funding research projects. DRAs also took multifaceted approaches to expand scientific collaborations through public–private partnerships, research funding mechanisms, and innovation networks. Cross-DRA communications were also reinforced through horizon scanning systems and consortiums to better inform and assist the regulatory decision-making process. The output measurements might be scientific publications, funded projects, DRAs interactions, and evaluation methods and guidelines. Improved regulatory efficiency and transparency leading to benefits to public health, patient outcomes, and translation of drug research and development as the key primary outcomes of RS development were anticipated but not yet clearly defined. Conclusion: The application of the implementation science framework is useful for conceptualizing and planning the development and adoption of RS for evidence-based regulatory decision-making. Continuous commitment to the RS development and regular review of the RS goals by the decision-makers are important for DRAs to meet the ever-changing scientific challenges in their regulatory decision-making process.