The results of computational fluid dynamics simulation for convective heat transfer of turbulent flow in a cooled convergent-divergent nozzle are reported. The importance of the heat transfer coefficient is to find the most suitable metals for the nozzle wall as well as its application for producing nano-particles. ansys-icem and ansys-cfx 13.0 are used to mesh and simulate fluid flow in the nozzle, respectively. Effects of grid resolution and different turbulence models on the heat transfer coefficient are investigated. Three turbulence models of k-�, k-ε and shear stress transport are applied to calculate the heat transfer coefficient. Stagnation absolute pressure and temperature are 10.3bara and 840K, respectively, the same as those in the experimental work. The heat transfer coefficients obtained from simulation are compared with the available experimental data in literature to find out the best suitable mesh grid and the turbulence model. Under the selected operating conditions, k-ε and k-� models have shown the best agreement with the experimental data with the average error of 6.5 and 10, respectively, while shear stress transport under predicts the values with 16 error. © 2013 Curtin University of Technology and John Wiley & Sons, Ltd. © 2013 Curtin University of Technology and John Wiley & Sons, Ltd.