A comparative analysis of the experimental data, which were obtained in a benchmark experiment on the thermohydraulics of a model assembly of fuel-element simulators in a flow of sodium-potassium alloy, and calculations performed by specialists, using thermohydraulic codes, from different countries is performed. The model assembly consisted of 25 fuel-element simulators arranged in a square. Russian specialists used the BRS-TVS.R code to perform calculations of the benchmark experiment, Japanese specialists used SPIRAL and AQUA, Spanish specialists used FLUENT, Dutch specialists used STAR-CD, and South Korean specialists used MATRA and CFX. The following experimental and computational parameters were compared: the coolant temperature in the channels under nonuniform geometric and thermal conditions in the assembly, the surface temperature of the measuring fuel-element simulator on the heated section, and the coolant velocity in the assembly cells around the measuring simulator. Special attention was given to investigating the influence of the spacing lattice on the coolant velocity.A benchmark experiment on hydraulics and heat transfer in assemblies of fuel-element simulators cooled by a liquid metal was performed as part of the conference of the international working group on the thermohydraulics of advanced nuclear reactors of the International Association of Hydraulic Research. The conference was held on July 5-9, 2004 in Obninsk. Specialists from the laboratories of the Physics and Power-Engineering Institute and the Scientific-Research and Design Institute of Electrical Technology prepared the specifications for the standard problem on the basis of experiments designed to validate the thermohydraulics of fast reactors with inherent safety [1][2][3][4][5][6][7][8][9].The goal of the benchmark experiment was to analyze the thermohydraulic characteristics of a model assembly, which is nonuniform with respect to geometry and heat release, in a flow of a sodium-potassium alloy (22% sodium + 78% potassium) with the simulators arranged in a square in the presence of a spacing lattice in regimes with variable energyrelease subzones and with an assessment of the reliability and accuracy of the computational results obtained using thermohydraulic codes. The following experimental and computational parameters had to be compared: the coolant temperature in channels with nonuniform geometric and thermal conditions in an assembly, the surface temperature of a measuring fuel-element simulator on the heated section with nonuniform geometric and thermal conditions in the assembly, and the coolant velocity in the assembly cells surrounding the measuring simulator.The standard problem was open, i.e., both the initial and boundary conditions necessary for performing calculations and the results of experimental investigations were presented in the specification, the latter in tabular and graphical form [10].