Incident swirling jets are complex flow, which involve a large variety of regimes. The numerical prediction of this type of flow can help understand the several phenomena present, which could allow development of more efficient processes, as well as assist in the evaluation of turbulence models. Aiming to evaluate the performance of different turbulence models in the predict of an axi-symmetric swirling jet impinging to a hot plate, the flow field was determined with the computational tool FLUENT, by employing two methodologies. The first one is based on the Reynolds average (RANS) and the second one is based on the Large Eddy Simulation (LES). Several two differential equations RANS models were tested, in order to identify their range of validity, by comparing with available experimental data. The methodology LES was evaluated with the dynamic Smagorinsky sub-grid model. In spite of the high cost, due to the need of utilizing a large memory space and simulation time, the results obtained with the LES methodology were significantly superior, due to the high anisotropy of the flow, which the RANS models have difficult to predict.