Two-phase critical non-flashing flow might occur, e.g. when a system filled with water and non- condensable gas as air or nitrogen, is emergency released. The common practice in Swedish industry is to employ one-dimensional system codes such as RELAP5 to calculate this type of flow. Such codes have in-build two-phase critical flow models which are intended for flashing steam-water mixture. However, non-flashing flows are different in its nature than flashing flows and because of this the default models might provide non-physical results. This paper aims to validate these models against experimental data and provide recommendations on how to predict non-flashing flows with satisfactory accuracy. Validation was performed against experimental data of  -  with water stagnation pressure conditions varying from 0.1 to 1.56 MPa, temperature subcooling in the range of 84.7 - 195.6 °C, maximum air mass flux of 383.2 kg/(m2 sec) and discharge section geometry arranged as an open pipe outlet, converging-diverging nozzle and converging nozzle.