When it is desired to model the behaviour of systems which depend on parameters that fluctuate through deterministic differential equations, we find some limitations when using these models in applications. For that reason we use stochastic coefficients (instead of deterministic functions) in the differential equations, achieving a better prediction in the variablility of the paramenters of the system. In this work we consider the problem of numerically approximating the solution to the wave equation posed over random media. Since the solution of this type of equations are stochastic processes, we use tools of probability theory such as the Karhunen-Loeve expansion to separate the deterministic part form the random part of the coefficients and solutions of the equations and then, we apply a finite element method in order to obtain an approximation on the main statistics of the solutions.