Discovery of new deposits has slowed dramatically since the beginning of the 21st century despite increased expenditure. Targeting below deep, complex cover is recognised as key to opening up new search-spaces and re-invigorating exploration success. Existing techniques and technologies are reaching maturity as direct-detection exploration methods are costly and have limited success when targeting beneath cover. Thus, the application of a predictive, conceptual-based approach to exploration targeting is required. The mineral systems concept provides a framework within which the critical elements of a system can be defined and incorporated into a predictive exploration targeting model. However, modelling of entire mineral systems remains in its infancy. As such, the effective translation of the mineral systems concept into robust targeting models is defined as a significant research challenge by several authors. A conceptual whole-system modelling approach is proposed in light of the critical elements of an orogenic gold mineralising system and a review of existing exploration tools. The approach combines probabilistic sub-models for predicting total mineral endowment with localised mineral distribution. Linking resource endowment assessments (total mineral endowment), with fluid dynamics and fluid-rock interaction modelling (local mineral distribution), provides an opportunity to develop resource-potential maps that define the approximate cost-benefit of exploring any given location. Further research is required for such modelling methods to become feasible. Although hydrodynamic and hydrogeologic modelling has advanced in other research fields, methods for modelling auriferous fluid flow through the crust remain under-developed. This, in part, is due to detailed 3D geological data being unavailable or prohibitively expensive to acquire. Potential solutions to limited data availability include the development of more advanced interpolation and geophysical inversion methodologies, based on data from well-understood geological provinces, resulting in improved 3D geological interpretations. Collaboration between academia and the minerals industry, to develop an integrated approach to entire mineral system modelling, may lead to successful predictive exploration, by providing a meaningful output to guide decision-making during the exploration targeting process.