Abstract: Existing sulfur-containing compounds, such as sulfate, sulfite and thiosulfate, can affect the determination of sulfide (H₂S, HS^- and S^2-) in geothermal water. In addition, the detection accuracy and precision may also be influenced by sample treatment processes such as sample collecting, sample pretreatment, and standard solution preparation. Zinc acetate and sodium hydroxide were added to water samples collected at the site in which sulfide would be precipitated as zinc sulfide and separated from water. The precipitation was further dissolved in hydrogen peroxide and Lefort aqua regia in which S^2- would be oxidized to SO₄^2-. The standard stock solution was prepared by using sodium sulfate, which was stable and easily purified. Sulfur content in water was determined by Inductively Coupled Plasma-Optical Emission Spectrometry (ICP-OES) using the spectrum of 182.624 nm as the sulfur element analysis line. It was found that S^2- concentration ranging between 0.1 and 100 mg/L was linearly correlated with emission intensity with a correlation coefficient of 0.9994. Detection limit of the method was 0.009 mg/L, relative standard deviation (n=11) was less than 1.80%, and standard addition recovery was 99.0%-103.0%. The technical indicators in this method are superior to previous analytical methods.