Successful exploration for hydrocarbon reservoirs requires understanding seismic response for different sandstone lithologies. Today, many new targets lie within shale-rich environments such as mud-rich deltaic systems, on continental slopes or other basins with shaly mobile substrates. To produce reliable predictions, any ambiguities created by changing sandstone reservoir properties must be reduced. As both lithology and pore fluids significantly influence seismic response, correctly modeling prospects requires methods which can discriminate between these separate effects. Evaluating amplitude variation with offset (AVO) permits identification of hydrocarbon reservoirs by their anomalous seismic response from background trends. This thesis tests empirical relationships between velocities and rock properties, effective media equations and Backus-averaging under several shaly-sandstone geometries ranging from pore-space clays to laminated sands and shales to successfully predict AVO response. Algorithms using the Hashin-Shtrikman equations for rock-strengthening pore-bridging clays and Backus-averaged laminae where shales reduce average acoustic impedance for the reservoir accurately modeled deepwater seismic data from the Gulf of Mexico. These two methods produce statistically significant models which reflect the seismic AVO response for varying sandstone properties