A numerical method, which is represented by both time- and volume-averaged transport analysis and direct numerical simulation of turbulence, was developed for thermal striping phenomena. The phenomena are characterized by a stationary random temperature fluctuation occurring in the region immediately above the fast breeder reactor (FBR) core due to a temperature difference of the core outlet coolant between subassemblies. The thermal striping phenomena are recognized as one of the key problems from the standpoint of high-cycle thermal fatigue of the in-vessel components such as the upper core structure, flow guide tube, etc. Fundamental experiments using water and sodium to simulate these thermal striping phenomena were calculated using the method developed in this study. Calculated results by the method were compared with the data under wide experimental conditions on the amplitude and frequency of the temperature fluctuations. Furthermore, the thermal striping phenomena in a 1:1 scale mock-up model sodium experiment simulating the outlet region of the FBR core were calculated by the method, and were compared with the calculational data. From these comparisons with the experimental data, it was confirmed that the numerical method has a sufficiently high potential in accuracy and efficiency to predict the amplitude and frequency of the temperature fluctuations related to the thermal striping phenomena. Consequently, it is concluded that the numerical prediction by the method developed in the present study can replace conventional experimental approaches using 1:1 or other scale model aiming at the simulation of the thermal striping phenomena in actual FBR plants. Furthermore, economical improvements in the FBR plants can be carried out based on the discussions of optimization and rationalization of the structural design using the numerical method.