Supersonic ejectors are widely used in a range of applications such as aerospace, propulsion and refrigeration. The primary interest of this study is to set up a reliable hydrodynamics model of a supersonic ejector, which may be extended to refrigeration applications. The first part of this work evaluated the performance of six well-known turbulence models for the study of supersonic ejectors. The validation concentrated on the shock location, shock strength and the average pressure recovery prediction. Axial pressure measurements with a capillary probe performed previously [Int. J. Turbo Jet Engines 19 (2002) 71; Conference Proc., 10th Int. Symp. Flow Visuzlization, Kyoto, Japan, 2002], were compared with numerical simulations while laser tomography pictures were used to evaluate the non-mixing length. The capillary probe has been included in the numerical model and the non-mixing length has been numerically evaluated by including an additional transport equation for a passive scalar, which acted as an ideal colorant in the flow. At this point, the results show that the k–omega–sst model agrees best with experiments. In the second part, the tested model was used to reproduce the different operation modes of a supersonic ejector, ranging from on-design point to off-design. In this respect, CFD turned out to be an efficient diagnosis tool of ejector analysis (mixing, flow separation), for design, and performance optimization (optimum entrainment and recompression ratios).