A wide range of stoichiometries has been previously proposed for soluble iron sulfide species and there is no general agreement on their importance in natural waters. The solubility of Fe(II) in 0.1 M NaClO4 equilibrated at 20 - 0.1°C with various partial pressures of H2S (0.1, 0.001, 0.0001, 0.00001 MPa) was measured in the pH range 3.1-7.9. Equilibrium was established within 1-6 h when amorphous FeS was the solid phase. The results could all be fitted using values for the solubility product constant (I = 0) of p*Ks = 3.00 - 0.12 and of the stability constant for a soluble Fe(HS) 2 species (I = 0) of p#2 = -6.45 - 0.12 where *Ks = aFe2+ · aHS-/aH+ and #2 = aFe(HS) 2/aFe2+ · (aHS-)2. Any soluble species of the form Fex (HS) 2x where x = &gif1; would fit the data equally well. Measurements at different partial pressures are inconsistent with labile species of the form FexSx. There was no evidence for a Fe(HS) + species. When a solution is saturated with respect to amorphous FeS, Fe(HS) 2 will only be a significant proportion of Fe(II) when S(-II) is higher than 0.2mmoll-1. The constants for Fe(HS) 2 or Fex (HS) 2x (x S 2) are consistent with all freshwater data where constant values of measured ion activity products provide no evidence for soluble complex formation. For marine waters with high sulfide concentrations (S 6mmol l-1), measured concentrations of Fe(II) are consistent with there being negligible soluble iron sulfide. The data are better fitted if the dissolved species are polymeric as predicted concentrations of the monomer Fe(HS) 2 are significant. These findings suggest that rather than the dissolved species being Fe(HS) 2, it is probably polymeric, that is Fex (HS) 2x (x S 2).