中药质量控制经历了由经验鉴别、成分检测向多成分及整体控制不断深化的过程。随着现代药品质量科学、分析技术、生产方式和数据环境发展，质量研究面临的核心问题正在由“能够检测什么”进一步转向“什么真正关系产品质量并需要得到保证”。以关键质量属性（critical quality attributes，CQA）为核心，从临床定位和产品理解出发，提出“药品质量概貌-候选质量属性-CQA辨识-物料、过程及中间物料研究-关键物料属性（critical material attributes，CMA）/关键过程参数（critical process parameters，CPP）辨识-质量形成规律-质量状态评价-生命周期持续确认”的方法学框架；强调CQA关键性应由具体产品质量风险和相应证据支持，CMA和CPP应基于CQA牵引的全过程研究辨识，全过程研究应服务于关键控制而非项目叠加；进一步讨论整体评价的非补偿性边界、可靠测量、回顾性与前瞻性研究及真实生产数据在持续确认中的作用。认为中药质量控制由成分导向转向质量属性导向，并非否定既有分析技术和质量标准体系，而是将研究起点由“发现什么、能够测什么”进一步转向“什么真正重要、为什么重要、如何形成及如何持续保证”。这一范式变革仍需通过不同类型中药产品和真实商业化生产数据持续验证和完善。

Quality control of traditional Chinese medicines (TCM) has progressively evolved from empirical identification and component testing toward multicomponent and holistic control. With advances in modern pharmaceutical quality science, analytical technologies, manufacturing methods, and data environment, the central question is shifting from “what can be measured” to “what truly matters to product quality and therefore needs to be assured.” Centered on critical quality attributes (CQA), this article proposes a methodological framework that starts from clinical positioning and product understanding and proceeds through “a product quality profile-candidate quality attributes-CQA identification-investigation of materials, processes, and intermediate materials-identification of critical material attributes (CMAs)/critical process parameters (CPPs)-quality formation laws-quality state evaluation-continuous lifecycle confirmation”. CQA criticality should be supported by product-specific quality risks and corresponding evidence. CMAs and CPPs should be identified through whole-process investigation guided by CQAs, and whole-process investigation should serve targeted control rather than expansion of control items. The article further discusses the non-compensatory boundary of holistic evaluation, reliable measurement, complementary retrospective and prospective studies, and the role of real-world manufacturing data in continual confirmation. A shift from a component-oriented to a quality attribute-oriented approach does not negate established analytical technologies or quality standards. Rather, it relocates the starting point of quality research from “what can be found and measured” to “what truly matters, why it matters, how it is formed, and how it can be continuously assured.” This paradigm transformation requires further validation and refinement across different types of traditional Chinese medicine products and real-world commercial manufacturing data.

R28