The Granites Gold Mine Callie gold deposit is located 550 km northwest of Alice Springs in the Tanami Desert. Gold mineralisation occurs due to a complex interplay of geological factors which include but are not limited to favourable lithology, structure, alteration and quartz veining. High-grade mineralisation, while pervasive under favourable conditions, is highly variable over short distances and is very nuggetty. In contrast, the lithology and geological structures that host the mineralisation are very continuous. Orebody modelling techniques have evolved over time because of a growing understanding of the mineralisation controls. Modelling issues such as sampling and assaying accuracy, modelling method, data density, gold population statistics, geological interpretation and domaining, grade boundary and calculation conditions, and model block size are constantly being reassessed in an effort to increase understanding of the mineralisation controls and modelling accuracy. The evolution of the modelling process in response to increasing data levels and a growing understanding of the mineralisation controls has resulted in changes to the perception of the orebody size and style. This has caused the choice of underground mining method to be revised several times and the process to final mining method selection to take longer than usual. In order to achieve the planned production start-up date of July 1998 the process has been allowed to progress in parallel with development of the underground mine. This paper describes the resource modelling and mining method selection process applied to the Callie underground mine and some of the issues which have contributed to its evolution. Some conclusions that are relevant to the efficient planning and development of new mining projects in general are presented.