Abstract: The appropriate amount of sewage sample was taken, and the pH was adjusted to 7‒8 with 1.0 mol·L⁻¹ hydrochloric acid solution or 1.0 mol·L⁻¹ sodium hydroxide solution for HLB solid phase extraction method, and was adjusted to less than 2 with hydrochloric acid for MCX solid phase extraction method. The solution was passed through the glass fiber filter membrane and polyether sulfone filter membrane in sequence. 200 mL of the above water sample was taken, and 200 μL of the mixed isotope internal standard solution was added. The solution was mixed well, and passed through solid phase extraction column activated with 6 mL of acetonitrile and 6 mL of water beforehand, controlling the column flow rate less than 4.0 mL·min⁻¹ or using natural gravity flow rate. After the sample was loaded, 6 mL of water, and 6 mL of 5% （volume fraction） methanol solution for HLB solid phase extraction method or methanol for MCX solid phase extraction method in sequence were used for rinsing. After continue pumping with the vacuum pump for 3 min, 6 mL of acetonitrile for HLB solid phase extraction method or acetonitrile solution containing 5% （volume fraction） ammonia water for MCX solid phase extraction method was used for eluting. The eluent was collected, and blown to dryness at 40‒45 ℃ by nitrogen. The residue was dissolved in 1.0 mL of initial mobile phase by ultrasonic oscillation and vortex, and the solution was centrifuged at 4 ℃ for 5 min. The supernatant was taken, and passed through 0.22 μm hydrophilic polytetrafluoroethylene filter membrane. The traditional drugs and fentanyl new drugs in the filtrate were determined by liquid chromatography-mass spectrometry, and quantified by isotope internal standard method. As shown by the results, linear relationships between mass concentrations of traditional drugs and fentanyl new drugs and peak area ratios of the target compounds to the corresponding isotopic internal standard were kept in definite ranges, with detection limits（3S/N） of 0.15‒1.0 ng·L⁻¹. Test for recovery was made by the standard addition method, with the relative recoveries of the two solid phase extraction methods in the range of 76.2%‒102%, and RSDs （n=6） of the determined values less than 9.0%. The two methods were applied to analysis of 50 actual sewage samples, and the detection rates were ranked from high to low as cotinine, hydroxycotinine, morphine, ketamine, codeine, and methamphetamine. After verification, the difference between the two pre-treatment methods based on HLB and MCX solid phase extraction was not significant, and the methodological parameters were similar. The HLB solid phase extraction method had good adsorption and retention effects on the target compounds, while MCX solid phase extraction method had stronger purification ability for sewage with higher impurity content. In practice, two extraction techniques could be used simultaneously to reduce determination errors in large volume water samples and decrease the false positive results.