Experiments and computational fluid dynamics/radiation heat transfer simulations of an 8 × 8 array of heated rods within an air-filled aluminum enclosure are performed. This configuration represents a region inside the channel of a boiling water reactor fuel assembly between two consecutive spacer plates. The rods are oriented horizontally or vertically to represent transport or storage conditions. The measured and simulated rod temperatures are compared for three different rod heat generation rates to assess the accuracy of the simulation technique. Simulations show that temperature gradients in the air are much steeper near the enclosure walls than they are near the center of the rod array. The measured temperatures of rods at symmetric locations are not identical, and the difference is larger for rods close to the wall than for those far from it. Small but uncontrolled deviations of the rod positions away from the design locations may cause these differences. The simulations reproduce the measured temperature profiles. For a total rod heat generation rate of 300 W, the maximum rod-to-enclosure temperature difference is 150°C. Linear regression shows that the simulations slightly but systematically overpredict the hotter rod temperatures but underpredict the cooler ones. For all rod locations, heat generation rates, and rod orientations, 95% of the simulated temperatures are within 11°C of the correlation values. For the hottest rods, which reside in the center of the domain where the air temperature gradients are small, 95% of the simulated temperatures are within 4.3°C of the correlation values. These results can be used to assess the accuracy of using simulations to design spent nuclear fuel transport and storage systems.