A water model experiment for thin slab casting was performed. The geometry of the water model was 1:2 under scaled in comparison to the size as operated in plant. The final goal of this experiment was to optimize the design of the submerged entry nozzles (SEN). Therefore, the SEN varied in the port gap from 6, 8, 10, 12, 15 to 18 mm. The mold width varied from 1150, 1400 to 1700 mm. Some other parameters were fixed constant, e.g. the mold depth was 112.5 mm, and the immersion depth of SEN in the mold was 210 mm. During experiment dye was injected, and a high resolution camera was used to record the flow pattern. The results of the water model experiment showed three different flow patterns: stable, transient and unstable flow. The gap size of the SEN was critical. It was evident that the results by using the SEN with the large gap, e.g. 18 mm, were always unstable. It was observed from the current experiment that unstable flow occurred when the SEN gap was larger than 10 mm. In order to understand the dynamics of the flow in the mold, a chosen case with the mold section of 1400*112.5 mm2 and a casting speed of 3.4 m/s was simulated. The simulation results were compared with the results of the water model experiment. Simulations with three different mesh sizes were evaluated. It was found that sufficient mesh refinement is required to reproduce the flow patterns as observed from the experiments.