Modeling the complex physical structures of cavitating flows makes numerical simulation far to be predictive, and still a challenging issue. Understanding the role of physical and parametric uncertainties in cavitating flows is of primary importance in order to obtain reliable numerical solutions. In this paper, the impact of various sources of uncertainty on the prediction of cavitating flows is analyzed by coupling a non-intrusive stochastic method with a cavitating CFD solver. The proposed analysis is applied to a Venturi tube, where experimental data concerning vapor formation are available in literature. Numerical solutions with their associated error bars are compared to the experimental curves displaying a large sensitivity to the uncertainties of inlet boundary conditions. Furthermore, this is confirmed by computing the hierarchy of most predominant uncertainties by means of an ANOVA analysis. Finally, a simple algorithm is proposed in order to provide an optimized set of parameters for the cavitation model, thus permitting to obtain a deterministic solution equal to the most probable one when considering physical inlet uncertainties.