Data derived from core and well-logs are essentially one-dimensional and determining eolian system type and likely dimensions and orientation of architectural elements present in subsurface eolian reservoir successions is typically not possible from direct observation alone. This is problematic because accurate predictions of the three-dimensional distribution of interdune and dune-plinth elements that commonly foam relatively low-permeability baffles to flow, of net:gross, and of the likely distribution of elements with common porosity-permeability properties at a variety of scales in eolian reservoirs is crucial for effective reservoir characterization. Direct measurement of a variety of parameters relating to aspects of the architecture of eolian elements preserved as ancient outcropping successions has enabled the establishment of a series of empirical relationships with which to make first-order predictions of a range of architectural parameters from subsurface successions that are not observable directly in core. In many preserved eolian dune successions, the distribution of primary lithofacies types tends to occur in a predictable manner for different types of dune sets, whereby the pattern of distribution of grain-flow, wind-ripple, and grain-fall strata can be related to set architecture, which itself can be related back to original bedform type. Detailed characterization of individual eolian dune sets and relationships between neighboring dune and interdune elements has been undertaken through outcrop studies of the Permian Cedar Mesa Sandstone and the Jurassic Navajo Sandstone in southern Utah. The style of transition between lithofacies types seen vertically in preserved sets, and therefore measurable in analogous core intervals, enables predictions to be made regarding the relationship between preserved set thickness, individual grain-flow thickness, original bedform dimensional properties (e.g., wavelength and height), the likely proportion of the original bedform that is preserved to form a set, the angle of climb of the system, and the likely along-crest variability of facies distributions in sets generated by the migration of sinuous-crested bedforms. A series of graphical models depict common facies arrangements in bedsets for a suite of dune types and these demonstrate inherent facies variability.