As observed by Worthington (2002), “The application of saturation-height functions forms part of the intersection of geologic, petrophysical, and reservoir engineering practices within integrated reservoir description.” It is also a critical reference point for mathematical petrophysics; the consequences of deterministic and statistical prediction models are finally evaluated in terms of how closely the estimates conform to physical laws. Saturations within a reservoir are controlled by buoyancy pressure applied to pore-throat size distributions and pore-body storage capacities within a rock unit that varies both laterally and vertically and may be subdivided into compartments that are not in pressure communication. Traditional lithostratigraphic methods describe reservoir architecture as correlative rock units, but the degree to which this partitioning matches flow units must be carefully evaluated to reconcile petrofacies with lithofacies. Stratigraphic correlation provides the fundamental reference framework for surfaces that define structure and isopach maps and usually represent principal reflection events in the seismic record. In some instances, there is a strong conformance between lithofacies and petrofacies, but all too commonly, this is not the case, and petrofacies must be partitioned and evaluated separately. Failure to do this may result in invalid volumetrics and reservoir models that are inadequate for fluid-flow characterization. A dynamic reservoir model must be history matched to the actual performance of the reservoir; this process often requires adjustments of petrophysical parameters to improve the reconciliation between the model’s performance and the history of production. Once established, the reservoir model provides many beneficial outcomes. At the largest scale, the model assesses the volumetrics of hydrocarbons in place. Within the reservoir, the model establishes any partitioning that may exist between compartments on the basis of pressure differences and, therefore, lack of communication. Lateral trends within the model trace changes in rock reservoir quality that control anticipated rates and types of fluids produced in development wells. Because the modeled fluids represent initial reservoir conditions, comparisons can be made between water saturations of the models and those calculated from logs in later wells, helping to ascertain sweep efficiency during production.