The primary target of the internal combustion engines design is to lower the fuel consumption and to enhance the combustion process quality, in order to reduce the raw emission levels without performances penalty.In this scenario the direct injection system plays a key role for both diesel and gasoline engines. The spray dynamic behaviour is crucial in defining the global and the local air index of the mixture, which in turns affects the combustion process development. At the same time it is widely recognized that the spray formation is influenced by numerous parameters, among which also the cavitation process inside every single hole of the injector nozzle. The proper prediction of the cavitation development inside the injector nozzle holes is crucial in predicting the liquid jet emerging from them. In this mechanism the CFD simulation is of great importance because of the too small dimension of the nozzle holes, which are mostly non suitable for an accurate experimental investigation and, when they are, these analyses need to be limited to a few cases for cost reasons. Nowadays the most used cavitation model is the two-phase homogeneous mixture model, especially with the Raleigh-Plesset closure model. In the literature it is possible to find out multiple examples of validation attempts of such a model versus experimental data: generally they fail because of the systematic overestimation of the mass flow with respect to the experimental data. The main parameters identified as responsible are the fluid properties, mostly in terms of the liquid density value, and the mesh setup.The focus of the present paper is to investigate the effective weight of each single fluid parameter, both for liquid and vapor phase, plus the mesh setup on the final pressure difference-mass flow rate curve, in terms of curve shape, curve values and evaluation of the cavitation onset/development.