The research investigated the cooling efficiency of a mist fan by redesigning its nozzle arrangement and determining the effects of nozzle arrangement on cooling performance, with emphasis on temperature distribution, velocity distribution and water-mass-fraction distribution. Five different nozzle arrangements were comparatively analyzed to determine the arrangement with the highest efficiency. A three-dimensional computational fluid dynamics (CFD) model based on a Lagrangian–Eulerian approach was used to simulate the interaction between the air and dispersed water droplets. The computational domain represented an enclosed room geometry measuring 7200 mm × 3600 mm × 3000 mm. A grid-sensitivity analysis was performed using three grids, and the reported temperature differences between successive grids were small. The five nozzle arrangements were compared at an air velocity of 5.96 m/s. The reported cooling efficiencies for arrangements 1, 2, 3, 4 and 5 were 66%, 67%, 68%, 70% and 63%, respectively, while the corresponding average temperatures were 304.2, 303.7, 303.6, 303.4 and 303.8 K. Nozzle arrangement 4 produced the highest reported cooling efficiency and the lowest average temperature. The temperature contours also indicated a more uniform distribution for arrangement 4. The further compared the CFD predictions with experimental measurements and reported strong agreement between the two approaches.