Depositional (facies) models of fluvial architecture permit straightforward categorization of deposits, but are necessarily simplistic. Here we describe a complementary database methodology which is designed to encapsulate the inherent complexity of fluvial systems and their preserved deposits. The database is implemented as a series of tables (characterizing qualitative and quantitative architectural and geomorphological properties and system attributes) populated with data derived from peer-reviewed studies of both modern rivers and ancient fluvial successions, and from other reliable sources. Architectural properties (geometries, internal organization, spatial distribution and reciprocal relationships of lithosomes) are assigned to three different orders of genetic bodies organized in a hierarchical framework, ultimately belonging to stratigraphic volumes that are homogeneous in terms of their controlling factors and internal parameters. Interrogation of the database generates a varied suite of quantitative information, whose principal applications include: (i) the quantitative comparison of fluvial architecture to evaluate the relative importance of intrinsic and extrinsic controls; (ii) development of quantitatively justified fluvial depositional models through the integration of data from multiple sources; (iii) development of better constraints on the workflows used to infer borehole correlations and to condition stochastic models of subsurface architecture; (iv) identification of appropriate modern and ancient analogues for hydrocarbon reservoirs.