The liquid bismuth collector model has previously been proposed to explain the common association between gold and bismuth seen in many ore deposits, and involves the scavenging of gold from hydrothermal fluid by liquid bismuth. Here, textural relationships and temperature and chemical conditions of mineralization at the Stormont Bi-Au skarn prospect in northwestern Tasmania are analyzed to investigate whether the liquid bismuth collector model can explain gold accumulation at this prospect. The calc-silicate skarn assemblage containing grandite, hedenbergite, actinolite, magnetite, epidote, quartz, and calcite suggests that initial metasomatism involved temperatures >400° to 500°C. Visible gold is associated with native bismuth, bismuthinite, maldonite, galena, and an unknown Bi-Te-S mineral. Native bismuth is directly associated with grandite, which formed during initial stages of metasomatism. This implies that bismuth would have precipitated from a hydrothermal fluid as a liquid. Heating experiments showed that bismuth alloy inclusions in garnet melt between temperatures of 250° and 288°C. The mineralization setting at Stormont and textural relationships suggest that gold was scavenged from hydrothermal fluids by liquid bismuth. Rhythmically zoned grandite crystals provide evidence for fluctuating hydrothermal fluid conditions, which may have contributed to dissolution and reprecipitation of gold, thus facilitating an ore zone refining process within the prospect. This refining process is made possible by the tendency of liquid bismuth to scavenge gold from significantly under-saturated fluids. Infiltrating gold-undersaturated fluids can dissolve gold not in contact with bismuth, which can then be preferentially redeposited upon encountering a bismuth droplet, enhancing the correlation between the two elements. A long-lived and focused fluid flow system would maximize the efficiency of this ore zone refining process, which is likely to be relevant to gold deposition at a range of deposit types.