Reliably predicting lithologic and saturation heterogeneities is a key problem in reservoir characterization. This study shows how near and far offset seismic impedance can be used to classify lithologies and pore fluids. The near offset seismic stack approximates a zero offset section, giving an estimate of the normal incidence acoustic impedance (AI=&rgr;V). The far offset stack gives an estimate of a VP/VS related elastic impedance (EI) attribute, equivalent to the acoustic impedance for non-normal incidence. These attributes can be computed from log data. Well data can be used, prior to seismic inversions, to test the feasibility of using AI-EI attributes for lithofacies identification. Bi-variate AI-EI probability distribution functions can be estimated from logs to obtain classification success rate for each facies. Success rate is a measure of the value of far-offset data for reservoir characterization. Examples are presented from North Sea, Gulf of Mexico, and Australia. ¿ 1998 American Geophysical Union