Queensland is similar to many regions of the world in having a diverse history of describing, mapping and classifying soils. Much of the 50-year catalogue of information on soil and land resources is stored in the Soil and Land Information (SALI) relational database system. This has enabled us to contribute effectively to the Australian Soil Resource Information System (ASRIS – www.asris.csiro.au). In the process, we needed to define a system that produced information relevant to diverse natural resource management issues from the primary data source. SALI has three basic types of information on soil and land resources: • Sites – observations of soil profiles and landscape attributes at geo-referenced points • Soil Profile Class (SPC) – allocations of soil profiles to local soil classes • Polygons – spatial and descriptive information for polygons with a large range of associated data. Data ranges from allocation to local soil class to each polygon is treated as a unique individual area and attributed as such. SALI contains approximately 84,000 sites, 3,200 Soil Profile Classes and 150,000 polygons. Approximately 80% of our land resource information is captured in the digital relational database. These data were used to estimate individual soil attributes for the idealised 5-layer soil that forms the primary data structure in ASRIS. Our system estimates 30 attributes for each of the 5 layers. McKenzie et al. (2005) provide definitions for the layers and attributes. The system we implemented was developed in an Oracle database using a series of modular procedures to populate the range of ASRIS data tables. This was achieved by writing some 12,000 lines of database code that will derive all the ASRIS attributes for sites, SPCs and polygons. Attribution of sites is from the raw data collected at each geo-referenced point using a series of rules to convert a complex site to an idealised 5-layer profile. Attribution of SPCs is achieved using the same code with the extra steps of estimating missing attributes from the representative sites recorded for that SPC. The attribution of polygons, in order of preference, is from representative sites, the SPC recorded, any site within the polygon and finally using a soil classification lookup table. The attribution of ‘all' polygons in Queensland has allowed the creation of spatial coverages to access the best available data at any point across the state and the appropriate level of detail. The main aim of this complicated attribution process is to retain local knowledge from land resource surveys rather than relying on a few high level pedotransfer functions for the state or a lookup table based on a continental soils map to provide the state-wide coverage. The use of modular database code to complete this process allows our data to be updated at anytime for inclusion in ASRIS with online access. Shortly after new data is collected, or refinements are made to polygons, the ASRIS website will display the alteration. With this system in place, the attribute surfaces are enhanced by the collection of primary data and data collection is informed by the system itself. With millions of pieces of data across some 300,000 features it is impossible to update, or even revise, any part of the system manually within an acceptable timeframe. The ability of the system to rapidly derive improved information products from foundation data and knowledge stored within soil and land resource information is the single biggest benefit of the approach taken. The continual improvements to information products critical to natural resource management projects will benefit all users of this data, especially modellers.