In this case study we compare the sampling quality of diamond drilling (DD) and reverse circulation (RC) drilling at the high sulfidation Purnama gold-silver (Au-Ag) deposit located in northern Sumatra, Indonesia.There was a systematic bias of 23 per cent gold metal under-call in the diamond drilling based Ore Reserve estimate compared to the mill reconciliation over three years of production. The RC based Grade Control data and mill production data were in much closer accord. Investigation led to quantification of the significantly reduced sampling imprecision (or improved precision) of RC drilling samples relative to diamond drilling samples and the determination of a grade bias between the drill types caused by sample volume and metal distribution in the mineralisation. An infill RC drilling program was undertaken to improve data density across the pit for inclusion in the Mineral Resource estimate. This subsequently increased Au metal in the estimate by 16 per cent.The sampling imprecision of primary diamond drill sampling is rarely tested. Duplicate sampling can be used to quantify sampling imprecision at each stage of subsampling. In diamond drilling it is tested at the primary sampling stage via second half core sampling, while RC sampling is tested with field duplicate sampling at the rig. Duplicate data can be compared in a number of ways to determine imprecision including Thompson-Howarth plots which quantify imprecision to allow comparison between different methods and protocols. Comparing sampling imprecision between data sets and data types allows determination of optimum drilling and sampling protocols and can reflect upon the underlying validity of Mineral Resource input data and estimation outputs.CITATION:Sims, D A, Pocoe, J, Kasnanto, A N and Pulungan, B, 2017. Why combine reverse circulation and diamond drilling in Mineral Resource estimation at Purnama?, in Proceedings Tenth International Mining Geology Conference 2017, pp 413-422 (The Australasian Institute of Mining and Metallurgy: Melbourne).