The purpose of the “International Wet Steam Modeling Project” is to review the ability of computational methods to predict condensing steam flows. The results of numerous wet-steam methods are compared with each other and with experimental data for several nozzle test cases. The spread of computed results is quite noticeable and the present paper endeavours to explain some of the reasons for this. Generally, however, the results confirm that reasonable agreement with experiment is obtained by using classical homogeneous nucleation theory corrected for non-isothermal effects, combined with Young’s droplet growth model. Some calibration of the latter is however required. The equation of state is also shown to have a significant impact on the location of the Wilson point, thus adding to the uncertainty surrounding the condensation theory. With respect to the validation of wet-steam models it is shown that some of the commonly used nozzle test cases have design deficiencies which are particularly apparent in the context of two- and three-dimensional computations. In particular, it is difficult to separate out condensation phenomena from boundary layer effects unless the nozzle geometry is carefully designed to provide near-one-dimensional flow.