The approach taken in this paper, namely synthesising a wealth of previous information with new data and a genetic model, in combination with integrated numerical analyses, led to new insights into the geological controls on the localisation of auriferous veins and residual prospectivity of the Charters Towers goldfield, NE Australia. The method also has implications for the assessment of other "mature" goldfields worldwide. Despite a number of different ore controls having operated within the Charters Towers goldfield, the controlling factors can be linked to a single genetic model for orogenic, granitoid-hosted lode-gold mineralisation in a brittle deformation regime (D4) of NE-SW to NNE-SSW shortening, under conditions of supralithostatic fluid pressure and low stress difference. Spatial autocorrelation results suggest district-scale alignment of the auriferous veins parallel to and overlapping with the ESE-WNW- to E-W-striking Charters Towers-Ravenswood lineament, a major crustal boundary in the basement to the Ravenswood batholith. At the camp-scale, auriferous veins have abundance and proximity relationships with NW-SE-, NNW-SSE-, NE-SW- and ENE-WSW-oriented lineaments, suggesting that structures that controlled gold deposition in one camp did not necessarily control mineralisation in other camps. Fractal dimensions obtained with the box-counting method range from 1.02 to 1.10, whereas veins in the Charters Towers City camp are characterised by a significantly higher fractal dimension of 1.28. This discrepancy may be taken to imply that most or all outcropping and near-surface deposits within the Charters Towers City camp have been found and that new discoveries are more likely to occur at greater levels of depth, or outside the boundaries of this camp. The new understanding has implications for the assessment of the residual prospectivity of the Charters Towers goldfield, where large areas of prospective rock types and structures (e.g., approximately 40% of the Charters Towers-Ravenswood lineament) are hidden under cover. This parameter space was inaccessible to the historic prospectors and has received relatively little attention from recent explorers. The following steps are suggested for the development of a targeting strategy for lode-gold exploration in areas of the goldfield under cover: (1) identify from geological and geophysical data the ENE-WSW (?? 15??) and NNW-SSE (?? 15??) striking structures and geological boundaries within a 20-km-wide corridor parallel to and centred upon the Charters Towers-Ravenswood lineament, the potential control on gold deposit distribution at the regional- to district-scale, (2) interpret from geological and geophysical data the distribution of pre-Middle Devonian granitoids within these areas that are the preferred host rocks of the payable gold deposits, (3) deduce from geophysical data the ENE-WSW (?? 15??) and NNW-SSE (?? 15??) striking structures that cut or bound the intrusions identified in step 2, (4) locate segments along the structures identified in step 3 that deviate most from the geometry of a straight line (e.g., potential bends or splays) and/or intersect other structures or geological contacts, or both, and (5) define and rank potential targets within the prospective areas identified in step 4 and systematically test the best ones. ?? 2006 Elsevier B.V. All rights reserved.