In this study a computational fluid dynamics (CFD) method has been developed to simulate the effect of pore morphology and its distribution in a 2D micromodel on the enhanced oil recovery factor of nanofluid flooding. Seven types of micromodel with different schematics and pore shapes were considered. SiO 2 nanoparticles, dispersed in distilled water, were used for the preparation of the nanofluid and flooding operation. To generate the desirable porous media, the geometries of the micromodels were generated using the commercial grid generation tool, Gambit 2.3. Then, the momentum and mass transport equations were solved based on the finite volume method using the Fluent 6.3 software to investigate the displacement of oil at the pore scale. In order to better understand the nanoparticles' effects and to confirm the validity of the CFD simulations, numerical results have been compared with the experimental data. The influences of some parameters such as heterogeneity of pores, connectivity of pores with or without throat line, tortuosity and pore shape on the enhanced oil recovery, breakthrough time and fluid trapping in the porous media were investigated. From the results, it has been found that random generation of pore distribution illustrates better results compared to homogeneous pore distribution. In addition, with the presence of nanoparticles in the injected fluid the number of fingers decreases. The fingering effect has the main effect on the oil recovery factor with a lower fingering effect having a higher recovery factor. So, in the homogeneous pattern the nanofluid flow in the porous media is uniform and symmetric. But in the random distribution model, the fluid flow is more realistic and similar to the fluid flow in reservoirs.