The current study aims to investigate the effect of nanoparticles on heavy oil viscosity and to recovery heavy oil from core sample by steam injection process. The cores being examined in the current research were categorized in two types involving: sandstone and carbonate (limestone). The cores were formed in cylindrical plugs, which their diameter was 3.8 cm and their length is 7.5 cm. The cores were prepared from the Iranian heavy/ medium oil reservoirs. The absolute permeabilities of the sandstone and carbonate samples were 170 mD and 0.1 mD respectively. The porosity ranges of the both samples types were between 18 and 30%. The heavy oil samples (13.25 and 21 API) were obtained from Iranian heavy oil fields. In the first phase of the current research, the effect of nanoparticles on heavy oil viscosity was studied. Various metal oxide nanoparticle, involving: Fe 2 O 3 , NiO, TiO 2 , WO3, CuO, ZnO and Al 2 O 3 were used in the experiments performed in the current research. The results showed that Fe 2 O 3 , NiO, TiO 2 , and WO 3 nanoparticles reduced the viscosity of heavy oil at low concentrations and high temperatures, while zinc oxide, alumina, and copper oxide nanoparticles increased the heavy oil viscosity in the same situation. Results showed that there were the optimum concentration and temperature for each nanoparticle in which the minimum viscosity of heavy oil takes place. The maximum viscosity reduction observed in 0.2 wt% concentration of Fe 2 O 3 nanoparticle and 100 ° C temperature was about 65%. In the second phase, the impact of the four effective nanoparticles on the steam injection process was investigated. In these experiments, the mixture of nanoparticles and distillated water vapor were used. Furthermore, the effect of the core permeability on heavy oil recovery was investigated. The experimental results showed that the injection of mixture of steam with some nanoparticles, such as iron oxide(III) and tungsten oxide, causes to increase heavy oil recovery. The increases occurred by iron oxide(III) and tungsten oxide were about 1.8 and 1.5 times, respectively. In the third phase, one- and three-dimensional models were proposed using mass, energy, and momentum balances. The oil phase was assumed to be contained 5 pseudo compositions and the gas phase contains 5 pseudo compositions and steam. The proposed models contain a set of partial differential and ordinary equations. The partial differential equations were discretized by the explicit finite difference method. In each model, the explicit method extracted nine differential equations and the equations were solved by the matrix solution for each grid and time simultaneously. The experimental results were compared with the one-dimensional model. The comparison, which its error was about 6.2%, presents that there is a good agreement between them. Moreover, the three-dimension model results were compared with the CMG simulation results, and the comparison, which its error was 7.82%, presents that there is a good agreement between them. The effective parameters, such as permeability, quality of steam, and pressure, were evaluated on the heavy oil recovery using the proposed three-dimensional model. In the fourth phase, the injection of mixture of steam and nanoparticles has been simulated by the CMG software, and the simulation results were compared with the experimental results. In the simulation, three changes, involving: injected fluid viscosity, penetration of the nanoparticles in the core, and possible reaction for reduction of the heavy oil viscosity, were configured. After the comparison of the simulation results with the experimental results, the injection of the mixture of nanoparticles and steam for the actual size of the reservoir was simulated. The results showed that the heavy oil recovery by steam injection and nanoparticles was about 2 times recovery by the steam injection without any nanoparticles.