The main objective of this work is to study the drug release behavior from inert matrix systems by using computer simulation. This study allowed us to propose a new statistical method to evaluate the drug percolation threshold as a function of the exposed surface area of the device. The matrix system was simulated as a simple cubic lattice. The sites of the lattice were randomly occupied at various drug-excipient ratios. By simulating a diffusive process, the drug was delivered from the matrix system. The obtained release profiles were fitted to two different models: near the excipient percolation threshold, the square root of the time was well fitted, whereas close to (but above) the drug percolation threshold, the power law described accurately the release data. A relationship between the initial drug load and the amount of drug trapped inside the matrix system at infinite time was found. This relationship was conveniently described by an error function. Percolation thresholds in the matrix systems were determined from the latter relationship by using a nonlinear regression method. The assessment of percolation thresholds depends on the exposed surface area of the matrix systems. Moreover, estimated percolation thresholds were in agreement with the predicted values stated in the percolation theory.