Day-to-day operating practices at existing operations need provide high quality data for decisions that will impact directly on the potential value of the available resource, in addition to meeting standards of quality expected of data used in feasibility studies that will be completed to obtain additional capital required to finance operational improvements. Drill hole location, down hole survey and assay data standards of accuracy and precision have received considerable attention in recent years. The authors propose that geological descriptions of both cored and non-cored drilling warrant at least equal attention, considering their fundamental role in establishing the framework used for geological modeling and resource estimation. The iron ore deposits of the Middleback Ranges have been a major source of iron ore for almost a century. High-grade hematite-dominant ore is currently mined by OneSteel from three open pit operations supplying ore to the company's integrated steelworks at Whyalla. A review of OneSteel's future iron ore resources, commenced in 2002, highlighted the need for more flexible and comprehensive management of all forms of geological data used in resource evaluation. Driving this program was the information needs of feasibility studies of the Iron Magnet deposit, located beneath the existing Iron Duke and Iron Duchess mines. Prior to the beginning of 2003, resource evaluation drilling data collected by OneSteel was managed using an in-house system based on industry standard relational database software. The system was demonstrably effective in ensuring the fundamental integrity of basic drilling data, but lacked necessary capabilities for the acquisition and management of geological descriptions, geotechnical and geophysical data obtained by down-hole logging. OneSteel's existing system for the management and interpretation of geological descriptions was restricted to storage of a single three digit code for each lithological description interval. The original logs for a large number of older drill holes could not be located, having been either poorly archived or culled from the database in the belief that they did not contribute to the understanding of the resources currently being mined. The three digit codes represented a combination of lithological and stratigraphic descriptions, including many subtle variations in iron ore colour and texture believed to be associated with variations in geochemistry, but subsequently demonstrated to be either incorrect or inconsistently applied. This resulted in the establishment of a plethora of codes for a relatively small number of distinct lithologies. Improvements implemented by OneSteel geologists addressed these shortcomings by ensuring the new data management system: facilitated the objective description of lithological characteristics; distinguished between obsevations and interpretations (eg lithology versus stratigraphy); promoted consistency between descriptions provided by different geologists; and helped to ensure data integrity by developing an audit trail for alterations made to data subsequent to its acquisition. The data used to develop interpretations of mineralisation distribution and characteristics are fundamental to the effective utilisation of a mineral resource: any mine's most important asset. Resource related risk is widely recognised as the principal source of technical risks affecting resource development projects, whether they are new mines or expansions of existing operations. A key lesson from this work is that most commercial or in-house' developed geological data management systems are able to handle basic data validation tasks effectively. Geologists, however, need to ensure that the data management system is appropriately structured to help promote the acquisition of objective and consistent data for interpretation.