Spacecrafts are usually expensive and can not be put into mass production, thus it is very practical and necessary to do research on fault detection and diagnosis on spacecrafts. The Livingstone software system, developed by NASA, employs a group of multi-level qualitative logistic models to describe the behavior of the system, and detect and diagnose faults by comparing model prediction values and sensed values of the sensors. This paper is aimed at explaining the usage of Livingstone software in the area of fault detection and diagnosis on spacecrafts. The principle and working process of Livingstone are elaborated in detail. A simulation experiment is performed on a simplified model of spacecraft propulsion system. The reasonable fault causes and probabilities are analyzed. The experimental results show that Livingstone is an effective and reliable tool for fault detection and diagnosis, which can find the faults within system rapidly and correctly.