The validation of computer codes related to the thermal-hydraulic analyses of nuclear reactors is a challenging undertaking because of the complexity of the phenomena involved, e.g., boiling, condensation, and mixing. In the frame of the ongoing validation of the best-estimate system code TRACE, the present paper focuses on the phenomena taking place during the quenching of the hot surface of the fuel rod simulator with cold water. Since TRACE describes the physical phenomena with empirical correlations derived from experiments, it is necessary to ensure that these correlations are valid if applied to similar experiments but different boundary conditions. By means of an uncertainty and sensitivity study, the influence of the empirical models and their associated uncertainties on selected output parameters is quantified and the parameters with the largest sensitivity are evaluated.