The prediction of multiphase flow in porous media is of great interest in several fields like Oil & Gas, 2 sequestration, geothermal energy, among others.Subsurface modeling is challenging due to the difficulties in predicting microscopic phenomena like Viscous Fingering that manifest at the macroscale in channellings, affecting displacement efficiency.Viscous Fingering is characterized for causing finger-like patterns that channels the phase with the lowest viscous resistance in the porous media.This phenomenon is the root cause of heavy oil production inefficiency, causing excessive water production, increasing costs, and energy consumption.This study proposes a numerical model based on the complete set of the Navier-Stokes equations in 3D to predict this natural phenomenon.The numerical model, developed using commercial CFD software, is based on an Eulerian-VOF solution to simulate immiscible multiphase flow in porous media in 3D.To verify the numerical model, it was compared against experimental data for oil recovery on three different viscosity ratios, ranging from 10 1 to 10 3 .The deviation against the experimental data was below 10%, achieving a successful prediction on the oil recovery while qualitatively predicting Viscous Fingering in 3D in great detail.Several differences between the studied viscosity ratios were found too.Not only regarding the oil recovery but also noticeable differences between the fingering dynamics were found too.The numerical solution proposed in this study could be helpful for scientists interested in subsurface modeling in great detail, where the prediction of multiphase flow in porous media is of great interest to improve subsurface processes.