Abstract: Driven by the development of next-generation information technologies and emerging operational paradigms, modern supply chain systems are characterized by highly stochastic demand, complex system structures, and incomplete data, which increasingly challenge the applicability of traditional operations management models that rely on precisely specified probability distributions in dynamic environments. The root cause of this issue lies in model uncertainty, namely, decision makers’ imperfect knowledge of the true stochastic evolution of the system. In recent years, distributionally robust optimization has emerged as an important approach to addressing model uncertainty. By constructing ambiguity sets of probability distributions and optimizing system performance under worst-case scenarios, this approach provides a unified analytical framework for robust decision making under incomplete information. However, existing studies have largely focused on static or discrete-time decision problems, while systematic investigations of continuous-time queueing systems, dynamic pricing, and multi-stage operational decisions remain relatively limited. Motivated by this gap, this paper presents a systematic review of research on dynamic distributionally robust optimization in supply chain and queueing systems under model uncertainty. From the perspective of the sources and representations of model uncertainty, we summarize the main modeling paradigms of distributionally robust optimization in continuous-time stochastic control, with particular emphasis on representative studies in heavy-traffic queueing systems, dynamic pricing, and inventory control. The survey reveals that model uncertainty is a central feature of dynamic supply chain and operations management decisions, and that distributionally robust optimization offers significant advantages in addressing this challenge. Nevertheless, its systematic application to continuous-time modeling, multi-product and multi-stage decision problems, and complex system structures remains an open area for further research. Finally, we identify common challenges in the existing literature and outline several promising directions for future research.