Abstract The aim of this paper is to review the numerical simulation methods now being used for dealing with fluid displacement in porous media. One of them is the invasion percolation method, IP, which applies where capillary forces are dominant with respect to viscous forces at the fluid interface. The diffusion-limited agregation method, DLA, consists in the construction of an aggregate of particles from a single particle, the seed, with the restrictions imposed by a probability field gradient satisfying the Laplace equation. Finally, the method of stochastic capillary networks takes advantage of the analogy between the displacement of a fluid in a porous medium and the circulation of the electric current in a purely resistive circuit. A flow problem in a porous medium is said to be solved when analytical expressions relating the pressure and the velocity fields are obtained. From this viewpoint the IP method and the DLA method do not provide this kind of information, although the latter describes the behaviour of the fluids at the interface correctly. The method of stochastic networks, on the other hand, do provide both pressure and velocity fields through the simulation process itself. These fields enable one to calculate flow rates at the inlet and oulet faces of the network as well as inside it, so making it possible to match numerical and experimental results. These methods are mainly applied to assisted and secondary oil recovery as well as in the contamination of aquifers. In particular, the method of stochastic networks is a valuable tool that constitutes a suitable complement to petrophysical and fluid mechanical laboratory measurements, which are of the utmost importance for the reservoir engineer.