Carbonate minerals commonly occur in coals of many ages and from a utilisation viewpoint can be deleterious. Several studies have been undertaken of the carbonates in the Permian coals of the Hunter Valley, Australia, but few studies use a multi-technique approach. For this study, a combined petrographic, geochemical and mineralogical approach was used to determine the distribution and residence of carbonate minerals in coal that had been intruded by a dyke. The dominant carbonate assemblages comprise primary siderite in inertinite-rich microlithotypes and secondary calcite(-ankerite-dolomite)-dawsonite in vitrinite-rich microlithotypes. The secondary carbonates were found in both the aureole of heated coal and also in an unheated mine-face sample. It is believed that the secondary carbonate minerals precipitated from magma-derived fluids percolating through the coal following the emplacement of the intrusions. The textures and distribution of the secondary carbonate minerals suggest that the temperature and pressure of the fluids may be just as important in developing fractures near dykes (particularly those that have multiple phases of geometries), cleat mineralogy and coal textures as direct heating from the intrusion. The partitioning of primary siderite with inertinite and secondary carbonates with vitrinite indicates that it can be reasonably expected that there would be a partitioning of minerals in various density fractions derived from float-sink tests and consequently a partitioning of elements with inertinite-rich fractions containing elevated Fe levels and vitrinite-rich fractions containing elevated Ca, Mg and Al. This partitioning has implications for the behaviour of the coal during washing and combustion, and the composition of combustion products. (c) 2006 Elsevier B.V. All rights reserved.