After three attempts in the last 6 years to fully characterize the reservoir in the Caballos Formation in the Puerto Colon Field, different approaches to explain the reservoir heterogeneity have been suggested, each of which involves a great deal of uncertainty. Of particular importance is the fact that the oil-water contact has been found at varying depths from well to well. This variation could not be satisfactorily explained structurally nor stratigraphically. Bouttete et al, 1993, on the basis of incomplete 2D seismic information integrated to an extensive set of geologic and engineering data stated that there are three different reservoirs in the Caballos Formation. In 1996, Calderon et al, using additional well and seismic data, concluded that there is a single tilted oil water contact in the Formation. There is strong evidence supporting each of these models but there are still some inconsistencies. Recently, Peña et al, performed a thorough structural, stratigraphic and petrophysical review. This latest approach considered improved data management procedures and geostatistical modelling in order to produce a better picture of the reservoir. A rigorous validation and integration process under a project database allowed the workteam to improve the understanding of the system under study. By combining geostatistical modeling with geologic expertise it was possible to reduce much of the uncertainty and choose the most probable static and dynamic models of the reservoir. It was found that there are not sufficient elements in the static and dynamic models to account for the existence of three different reservoirs. A tilted OWC was accepted as the best representation of the reservoir for the available information.