With few exceptions, the world's antimony resources are derived from hypogene deposits of stibnite which form mainly at temperatures between 150 degrees and 300 degrees C. In this paper, we investigate the stability relationships and solubility of minerals in the system Fe-Sb-S-O, in order to understand why stibnite is so dominant and to explain the distribution of the other minerals in hypogene antimony ore deposits. Textural relationships, chemographic analysis, and thermodynamic data are used to develop calibrated petrogenetic grids for this system. These grids indicate that stibnite is the stable antimony phase for a wide range of physicochemical conditions encountered in nature.At temperatures >350 degrees C typical crustal fluids require concentrations of thousands of parts per million Sb (an unlikely occurrence) in order to saturate with stibnite, whereas at temperatures <250 degrees C stibnite may precipitate from fluids containing as little as 1 ppm Sb. This observation, and the fact that stibnite deposits commonly have complex parageneses in which stibnite is relatively late, suggests that decreasing temperature may be an important control of mineralization. Many stibnite deposits, however, have relatively simple parageneses, and a significant proportion of these is hosted by black shales. It is thus possible that, in many cases, reduction was the principal cause of mineralization. This is consistent with the finding that antimony mineral solubility decreases precipitously with f (sub O 2 ) across the kermesite-stibnite boundary for a wide range of pH (< or = neutrality). Other stibnite deposits with relatively simple parageneses (including some of the world's largest deposits) have formed from solutions which apparently equilibrated with limestone and only became saturated with stibnite when they encountered shale. Intense pyritization is an important feature of some of these deposits. At the near-neutral pH conditions and high Sigma a s implied by equilibration with limestone and pyritization, respectively, Sb is transported largely as the species HSb 2 S (super -) 4 , and stibnite deposition is favored by decreasing pH and/or a (sub H 2 S) , both of which are predicted consequences of pyritization.Gudmundite and native antimony only display the ranges of solubility required to form economic deposits at f (sub O 2 ) , pH, and Sigma a s conditions that rarely occur in nature; these minerals are most effectively deposited by decreasing f (sub O 2 ) . Berthierite is only stable over extremely narrow intervals of f (sub O 2 ) , and f (sub S 2 ) , which probably explains why it is seldom found in large concentrations. Hypogene kermesite is also uncommon for similar reasons. At low temperature it replaces stibnite as the stable phase over a broad range of f (sub O 2 ) -pH conditions and consequently is a common supergene mineral. Senarmontite stability is restricted to extremely high f (sub O 2 ) conditions where its solubility is too high to permit hypogene saturation, except at extraordinarily high Sb concentrations such as might occur during remobilization of preexisting antimony mineralization.The common occurrence of high concentrations of gold in stibnite deposits and vice versa may reflect the fact that at pH conditions coinciding with the H 2 S-HS predominance boundary and f (sub O 2 ) conditions marginally below the sulfate predominance field, both Au and Sb can be transported in appreciable concentrations as bisulfide complexes. Deposition of native gold and stibnite is favored by decreasing pH, with or without reduction. It is probably no coincidence that deposits enriched in Sb and Au occur in settings which would have been capable of providing the relatively alkaline conditions necessary for transport of these metals and are hosted by rocks which were potentially able to cause the acidification needed to induce stibnite-gold mineral precipitation.