SUMMARYThe main objective of this work is to present a strategy for integrating seismic information and, potentially, other types of geophysical measurements in seismically-derived reservoir models of rock facies discriminated, e.g., by porosity, mineralogy, and fluid. Even though seismic data do not directly provide information for reservoir properties, such as facies distribution, porosity and net-to-gross, these data can be used to quantitatively constrain reservoir models away from well control. We illustrate here two techniques for estimating facies distribution and rock properties from seismic amplitudes following a probabilistic approach and integrating them into reservoir models via geostatistical methods. This use of geostatistics allows us to obtain several realizations of high resolution models of facies and reservoir properties. We then show how to combine the two techniques to obtain fine-scaled models which honor seismic data. The resulting integrated method represents a closed loop strategy for reservoir modeling and it can be extended to incorporate various types of data.