For the investigation of fluid flow in the supercritical state, computational fluid dynamics (CFD) have become increasingly important to gain information that cannot be obtained experimentally. Especially at near-critical pressures, the application of CFD methods is still a challenging task due to the strongly varying thermo-physical properties that complicate the solution process of the Favre-averaged conservation equations. The focus of the present work is on a numerical model (ANSYS FLUENT (R)) simulating flow and temperature fields of supercritical water jets discharged in a subcritical water bath at near-critical pressures. However, the used turbulence model tended to overestimate the values of the turbulent thermal conductivity near the pseudo-critical point of water. Various approaches based on a variable turbulent Prandtl number are presented to deal with these difficulties. Additional experiments were performed in a high-pressure vessel to record axial temperature profiles of the supercritical water jet and validate the numerical model. (c) 2012 Elsevier B.V. All rights reserved.