Ore characterisation is an integral part of mineral processing especially in the gold mining industry where the use of an optimised process can mean the difference between economic and uneconomic ventures. In this study a method has been proposed using a combination of automated image analysis in the form of QEMSCAN and bulk chemical techniques in the form of diagnostic leaching. These methods can be used to characterise the 'visible' gold occurrences in complex gold ores as a precursor to more in depth analysis and identification of 'invisible' gold by microbeam analysis. It has been shown that by using this combined method it is possible to accurately identify both the distribution and associations of gold by QEMSCAN, with the results of diagnostic leaching used to verify this data and ensure a representative sample had been selected. (C) 2005 Elsevier Ltd. All rights reserved.