Results of gold determination by fire assay method from different laboratories are compared and discussed. Marked discrepancies in Au analysis particularly at low levels were noted, The discrepancies were observed to be due to the different sample weights used, matrix composi-
tion and sampling error. It was found that acceptable precision and accuracy in fire assaying for Au and Ag at low concentration can be obtained with at least 50 gm sample weight ground to -200 mesh. It was also observed that analysis of the gold prill produced by fire assay and
determined atomic absorption is a better procedure than straight fire assay. Straight atomic absorption analysis of rock and soil samples can be utilized as a reconnaissance method of isolating targets but not to base gold credits on.