Abstract: In this research, the possibility of gold content in gold products manufactured and distributed fall short of standards due to inadequate analysis has been studied. Total 25 samples (5 samples each sold as 24K gold, 18K yellow gold and white gold, and 14K yellow gold and white gold) were prepared, and the gold content in each sample was analyzed by two different methods:XRF (5 runs per sample) widely used in manufacturing industry and cupellation designated as standard procedure usually. In 24K gold sample, the difference between two analytic methods for gold content above 999‰ was determined as 1.23‰, which is relatively small. However, the difference in Au990 products became greater and was found to be 4.5‰. In 18K gold samples, the difference became even greater, and difference for yellow gold and white gold were 6.4‰ and 13.1‰, respectively. Likewise, differences in 14K yellow gold and white gold were 18.2‰ and 16.5‰, respectively. From above results, it can be concluded that difference in XRF analysis becomes greater when gold content is smaller and that data scattering in white gold that contains nickel alloy element is greater than that in yellow gold. Therefore, XRF should only be used for gold content above 999‰ for qualitatively analysis of gold content, and results on products with gold content below 999‰ are not reliable. We conclude that cupellation is a more suited method for quality control of gold content than XRF.