Abstract: Ten portions of water samples taken from the water source, water works and pipeline network, were poured fully and separately into 2 L-brown glass bottles, and 100 mg of Na₂SO₃ were added to each bottle of water sample to expell remained Cl₂. The bottled water samples were kept at 4℃ and should be used for analysis within 14 d. 100 μL of 20 mg·L^-1 mixed internal standard solution of acenaphthene-d₁₀ and chrysene-d₁₂ were added to 1 L each of the water samples, which was then treated by solid phase extraction as described in the national standard of GB/T 5750-2006. The eluate obtained from solid phase extraction was dehydrated, and evaporated to near dryness, and the residue was taken up and made up to 1.0 mL with acetone. The acetone solution was separated by gas chromatography on HP-SMS chromatographic column (30 m×0.25 mm, 0.25 μm). Separation of the 23 semi-volatile organic compounds in the acetone solution was performed under the mode of programed temperature elevation in the temperature interval from 55℃ to 310℃, and determined by MS adopting EI and SIM. Linearity ranges for the semi-volatile organic compounds were found in the same interval between 0.20-10 mg·L^-1, except for pentachlorophenol was in the interval between 1.00-50 mg·L^-1. Their detection limits (3s) were found in the range of 0.03-0.18 μg·L^-1. Recovery was tested by standard addition method using a negative sample as matrix. Values of recovery were found in the range from 75.5% to 124%. Values of RSDs (n=6) were found between 1.1% to 16%. None of the 23 semi-volatile organic compounds was found in the 10 water samples, proving that the quality of city water meets to the requirements of national standards relating to limitations of the semi-volatile organic compounds.