<p>Digital twin (DT), defined as a high-fidelity virtual representation that dynamically synchronizes with its physical counterpart through real-time data integration, offers a transformative solution for tool condition monitoring (TCM) in the milling of thin-walled parts. Thin-walled parts demand rigorous real-time monitoring of tool conditions because of their low structural rigidity and high susceptibility to deformation during machining. Traditional tool wear monitoring (TWM) methods have problems such as hysteresis and reliance on empirical thresholds, while DT can realize high-precision and real-time TCM and prediction by constructing a bidirectional dynamic mapping between the physical tool and its virtual model. Specifically, this technology integrates multi-source sensor data, such as cutting forces, vibration, and acoustic emissions, during machining in real time. Combined with machine learning algorithms, it assesses tool wear and predicts remaining tool life in real time, thereby optimizing tool change strategies. More importantly, given the proneness to deformation of thin-walled parts, DT can predict machining deformation through virtual simulation and reversely optimize tool paths and process parameters, enabling active control and compensation of machining errors, thereby ensuring the machining accuracy and efficiency of thin-walled parts. This paper reviews the application and research progress of DT in milling TCM for thin-walled parts. The key technologies, realization paths and future challenges are analyzed. Firstly, the classification of thin-walled parts and machining characteristics are introduced. It is crucial to monitor the condition of the tools in the milling process to guarantee the precision and efficiency of machining. Next, the development of DT technology and its applications in milling are discussed. The DT technology can enhance the accuracy and reliability of TCM. The methods involved in TCM are examined, including signal acquisition and preprocessing, feature extraction, and decision-making processes. Relevant examples are provided to illustrate the relationship between thin-walled part machining and TCM. Finally, the limitations of the current DT technology are highlighted, alongside future challenges. These challenges include issues such as model accuracy being constrained by data quality, real-time limitations due to computing power, and latency problems related to virtual-real synchronization. The DT technology provides an intelligent tool for efficient precision machining of thin-walled parts, but its full-scale implementation still requires interdisciplinary collaboration and technological breakthroughs. This review can be used as a reference for the related research and engineering applications.</p>