Abstract: The TOPAZ-Ⅱ thermionic space nuclear reactor, with its compact structure, high power density, and efficient thermionic energy conversion, is an ideal candidate for space-based power supply in deep-space exploration and long-duration missions. To ensure its safe and stable operation under both nominal and off-nominal conditions, a transient analysis methodology for the TOPAZ-Ⅱ thermionic space nuclear reactor system was established based on a modified version of the RELAP5 system code. The method fully incorporated the primary design features of thermionic space nuclear reactors, including detailed physical models of the thermionic fuel elements (TFE), reactor core, NaK-78 coolant loop, and the space-adapted loop radiator system. Temperature reactivity feedback effects of key materials, such as uranium dioxide fuel, molybdenum emitter and collector electrodes, and the zirconium hydride moderator, were explicitly modeled to capture the reactor’s dynamic response characteristics. The developed RELAP5 code was validated through two complementary approaches. First, steady-state simulation results were compared with nominal design parameters of the TOPAZ-Ⅱ reactor. Relative deviations in key thermal-hydraulic indicators, such as core power and coolant inlet/outlet temperatures, were all within 1%, confirming the model’s accuracy and consistency with established reference values. Second, the model’s ability to simulate transient responses was verified against experimental data from the V-71 electrically heated non-nuclear test facility. Under both partial loss-of-flow and power ramp conditions, the calculated temperature response trends matched the test data well, validating the model’s effectiveness in capturing system dynamics. Based on the validated model, two representative accident scenarios were further investigated: A reactivity insertion accident (RIA) caused by unintended rotation of the control drum, and a loss-of-coolant accident (LOCA) triggered by a hypothetical sudden breach. These scenarios reflect critical failure modes under deep-space mission conditions. Simulation results indicate that although the system exhibits positive temperature reactivity feedback, mainly due to the ZrH moderator, the reactor maintains thermal safety margins throughout all transient phases. Specifically, in the most severe RIA case (0.01 insertion), peak fuel and emitter temperatures reach approximately 2 600 K and 2 100 K, respectively. For the LOCA scenario caused by complete coolant loss, three shutdown strategies were compared. Under the worst-case condition, peak fuel, emitter, and stainless steel cladding temperatures reach 2 500, 2 150, and 1 600 K, all remaining below design limits. In all cases, the reactor is safely shut down before thermal limits are exceeded, verifying the model’s ability to capture nonlinear feedback and shutdown margin. Overall, this work establishes a validated and physically consistent simulation methodology for thermionic space nuclear systems. The framework supports further safety evaluation, mission planning, and digital model development for future deep-space reactor applications operating under extreme environmental conditions.