The complex geological conditions in tunnels pose a huge challenge to the reliability of tunnel boring machine (TBM). However, existing reliability studies typically focus on core structures such as cutters and cutterheads, with less consideration given to the rest of the components that frequently fail. In this study, the reliability analysis and dynamic evaluation of TBM components with high failure rates are carried out relying on the Shanxi Central Yellow River Diversion Project. The life distribution and reliability variation characteristics of TBM components under different rock mass classes are investigated in terms of tunnelling time and tunnelling distance as two types of life data indexes. And the life index which is more suitable for the reliability evaluation of TBM components is identified by comparison. On this basis, a dynamic evaluation method for the reliability of TBM components under the condition of multi‐classes surrounding rock is proposed. This method can quickly evaluate the current reliability of TBM components and serve as the basis for preventive maintenance. The results of this study play a certain role in supplementing the reliability research of TBM and also provide a scientific basis for optimizing the design and maintenance strategy of TBM components.