Water injection is a common method to maintain reservoir pressure and improve oil recovery. The efficiency of oil recovery in the case of heavy oils is limited by the high mobility ratio between the injected water and oil. As the world thirst for energy is increasing every year while oil production from known oil reservoirs together with the discovery of new oil reservoirs deplete at considerable rate, it becomes indispensable to use more effective methods to produce oil from known reservoirs. The injection of dispersions, in particular of oil-in-water emulsions, as an agent of mobility control of injected fluid has been tested and studied with relative success. However this technique is not completely developed and understood. The effective use of emulsion injection as an alternative for oil recovery needs a complete analysis of different regimes of emulsion flow through the pore space of a reservoir. If the drop size of the dispersed phase is of the same order of magnitude of the pore size or lager, the drops can agglomerate and partially block the flow through the pores, thus controlling the displacing fluid mobility, getting a more uniform displacing front and an increase in the oil recovery factor. The main goal of this work is the study of oil displacement process through a porous media by water and oil-in-water emulsion injection. Different experiments were carried out for analysis of different aspects of the problem, including the alternating injection of water and oil-in-water emulsion at different flow rates, through cores with different permeabilities connected in parallel, and visualization of flow through a transparent non consolidated porous media, formed by glass beads. A model of emulsion flow was considered by modifying the relative permeability curve of the aqueous phase, which is written as a function not only of the aqueous phase saturation, but also as a function of the emulsion drop concentration and local capillarity number. The process of oil displacement by alternated water-emulsion injection was also studied numerically by a code developed in Matlab using TPFA (Two Flux Approximation) and IMPES (IMplicit Pressure and Explicit Pressure Saturation) methods.