At the end of preliminary hydrocarbon production by natural depletion, substantial amounts of hydrocarbon remain entrapped in reservoirs due to their heterogeneity and capillary forces. Knowing the amount of the trapped phase and mechanisms of trapping are crucial in proper designing of oil recovery projects. In this work, we evaluated the different trapping models that represent the relationship between initial and residual saturations and proposed a suitable model for different types of rock and fluids. The accuracy of each model was evaluated by different experimental data, and the model parameters were fitted by curve fitting. Results showed that the Spiteri et al. and Ma and Youngren models could predict residual saturations better than other trapping models. The maximum R2's of the Spiteri et al. and Ma and Youngren models were 1 and 0.99 for oil/brine systems, and 0.95 and 0.97 for gas-brine fluid pairs, respectively, regardless of the type of porous media. These two models accurately fitted the Initial-Residual (IR) experimental data of the gas-oil system that were obtained in this study. The effect of various parameters such as initial oil saturation, porosity, interfacial tension, flow rates, capillary numbers, and porous medium types on trapping in an unconsolidated sand pack were experimentally studied. The results imply that in the water-wet sand pack, the residual saturation increased by increasing the initial saturation and interfacial tension. However, increasing the porosity, mobility of oil phase, flow rate, and capillary number seems to have made for a decrease in the amount of trapped phase. It was also shown that trapping was more likely to occur in unconsolidated sand packs than in carbonate ones.