The CONTAIN code was used to predict the helium concentrations, gas temperatures and pressures, and wall temperatures of four experiments performed in the NUPEC 1/4-scale model containment. These experiments investigated the thermal-hydraulic effects of helium and steam source flow rates, source elevation, and internal water sprays. Two CONTAIN flow solvers and two nodalization schemes were assessed. One NUPEC test, International Standard Problem 35, was investigated in detail, including the pretest heating phase. The thermal hydraulics of this test were dominated by internal water sprays. A modeling approach based on the assumption that the water sprays generated a large air vortex yielded the best results. Reasons for deviations between the predictions and data are suggested based on experimental uncertainties, different analysis methods, and nodalization schemes.