This paper is dealing with modeling a heterogeneous reservoir rock from a Mexican offshore field. The problem is porosity characterization through the use of computed tomography images. The addressed issues include: association of secondary porosity with vugs and fractures, vug connectivity and spatial distribution, porosity associated to the rock, and vugs stochastic imaging. Porosity was obtained from computed tomography images, and calibrated with core data. From these images, exhaustive porosity variograms were obtained, showing a correlation range of about 20 millimeters. Further, use of indicator variables allowed vugs visualization, confirming the assumption that dissolution was the dominant process in vugs genesis. It is showed also that permeability has been enhanced through the connection of vugs by high porosity footprints. This detailed information has been used to reproduce geometry and spatial distribution of vugs through a sequential indicator algorithm. The stochastic images reproduce well vugs spatial distribution, explaining the secondary porosity role in the production mechanism in these reservoirs.