A novel methodology based on a statistical non-parametric approach is presented in this paper. It can yield more objective quantification of code physical model uncertainty by making use of model performance information obtained from assessment studies of appropriate separate effect-tests. Uncertainties are quantified in the form of estimated probability density functions (pdfs) calculated with a newly developed non-parametric estimator, and, by applying a novel multi-dimensional clustering technique, the methodology takes into account the dependency of a model's uncertainty on system conditions. These uncertainties are the objective input information needed by code uncertainty propagation methodologies applied for assessing the accuracy of best estimate codes in nuclear systems analysis. The new methodology has been applied to the quantification of the uncertainty in the RETRAN-3D void prediction model and then used in the analysis of a double LOCA transient in an integral-test facility. This has clearly demonstrated the basic feasibility of the approach, as well as its advantages in yielding narrower uncertainty bands for quantifying the uncertainty in the code's prediction of the void fraction evolution during the transient.