Several soil samples from different areas of China were extracted for plant available sulfur (S) with four extractants, CaCl₂、Ca(H₂PO₄)₂ (pH=4.0)、KH₂PO₄、and NaHCO₃ (pH=8.5). The total and available S in soils, and S contents in winter wheat in it's different growth stages were determined. The results showed that the available S in the soilwas significantly related to the contents of S in plants in all of four treatments. The results of using the inductively coupled plasma emission spectrometry (ICP-AES) and turbidimetry to determine soil available S showed extremely significant correlationship between the methods with correction coefficient being 0.867**. The contents of available S in soil determined by these two methods also had a significant correlationship with the contents of S in wheat, with correlation coefficients being 0.794**and 0.684**, respectively, indicating that both turbidimetry and ICP-AES were applicable to determination of available S in soils. But ICP-AES was better than the turbidimetry in stability. Using the Ca(H₂PO₄)₂ as extractant, the average variation coefficient of the results was 5.3%, the least in the four extractants, while using NaHCO₃, 8.3%, the largest.. Based on the characters of the extractants, the correlation of available S in soils with the S contents in wheat and the variation coefficient, it could be concluded that the Ca(H₂PO₄)₂ extractant and ICP-AES testing method were the best ones to determine the available S in soils.