Abstract: As the core component of loader working device, its fatigue life and reliability directly affect the performance and safety of the whole machine. For large structural components, due to their high testing costs and the limited number of available samples, traditional reliability assessment methods are often difficult to apply directly. Based on the fatigue test data of loader working device, three reliability evaluation methods of loader working device under very small sample are summarized, namely virtual augmented-sample & modified Bootstrap method, GM grey model & Bootstrap method and BP neural network & Bootstrap method. This study opens up a new path for the reliability evaluation method of expensive large-scale structural parts.