Abstract: The safety injection (SI) system might be put into use under the loss of coolant accident (LOCA) in nuclear reactor and the thermo-hydraulic phenomena while the subcooled water is mixed with pure steam in high temperature will directly affect the judgement on reflooding of core. For enhancing core competitiveness of nuclear power research and development design, China General Nuclear Power Group (CGN) independently developed and designed a system code of the hydraulic system of two-phase flow analysis named LOCUST, and the current version is LOCUST 1.2. The LOCUST code is written in C language and uses two-fluid model as the conservation equations, treats the gas and liquid phases as separate single phases, and calculates the interphase interactions. This code is generally used for the safety analysis of accident conditions such as LOCA. Nuclear safety regulations require that computer codes used in the safety analysis of nuclear power plants must be fully verified and validated. The use of experimental data is essential for the validation of the safety analysis code for the thermal-hydraulic system. As an important validation for LOCUST, the thermal mixing test of emergency core cooling (ECC) system based on the ECCS-XJTU test facility in Xi’an Jiaotong University was introduced in this paper. Based on this test facility, the ECC thermal mixing tests were conducted, which were mainly focused on the mixing between subcooled water injected from the SI pipe with a range of 25 to 125 kg/h in mass flow and the pure steam in the primary pipe with a range of 100 to 500 kg/h in mass flow. Besides, the modeling, calculations for ECC tests by using LOCUST code and analysis for calculation results and experimental results were performed. The experimental results show that the temperature stratification in the vertical direction is observed in fluid field after mixing, and the main temperature fluctuations concentrated at the lower part in the primary pipe. This is mainly due to the hydraulic jump phenomenon and the effect of gravity. The temperature of the measurement point at the lowest location (point Ⅳ at cross-section with six measurement points, point Ⅱ at cross-section with two measurement points) is much less than the saturated steam temperature, which could be regarded as the fluid temperature and compared with the fluid temperature calculated in LOCUST. The comparison between tests and simulations shows that the maximum relative error of LOCUST in mass flow of liquid is within 13.8%, and that in temperature is within 8%, which validates the reliability and accuracy in simulations of LOCUST for two-phase thermal mixing in LOCA.