Organophosphorus pesticides are one of the most widely used insecticides, which are mainly used in grain, vegetables and fruits. Methyl parathion is a kind of organophosphorus pesticide, which belongs to nerve agent. It can cause different degrees of poisoning to human and livestock and cause serious environmental pollution. Therefore, it is of great significance to establish an effective method for detecting methyl parathion residues in agricultural products. The determination of methyl parathion is often carried out by gas chromatography, but because of the strong polarity and thermal instability of methyl parathion, gas chromatography brings certain difficulty. The High Performance Liquid Chromatography (HPLC) method was researched for the testing of Methyl Parathion residues in vegetables, and the chromatographic conditions for sample extraction, purification and detection were screened and optimized. In vegetables, the matrix is complex, the pesticide residue is low, and there are many interference factors. The main residual components are not easy to separate, enrich and purify, so the detection of related pesticides is not accurate. After lots of experimental exploration, the chromatographic conditions by acetonitrile extraction agent, methanol:water (73:27) as mobile phase and UV detection wavelength choosing 270 nm was selected finally. What’s more, QuEChERS (Quickly, Easy, Cheap, Effective, Rugged, Safe) method, a new pretreatment technology for pesticide residue detection in agricultural products developed in the world recent years, was used for pretreatment of three kinds of vegetables, and PSA and GCB were selected as purifying agents for sample pretreatment ultimately. The experimental results displayed that the chromatographic peak area of Methyl Parathion exhibiting a good linear relationship with its concentration in the 0.05 μM~20 μM range, and the standard curve equation is Y=4833.5x-32.64, the correlation coefficient is 99.96%. The average recoveries of Methyl Parathion in three kinds of vegetables (Lettuce, Cucumber and tomato) were between 87.38% and 114.12% at the three spiked levels of 0.5, 2 and 8μM, and the relative standard deviation (RSD) was between 1.72% and 6.2%. This method has the good points of simple operation, accurate and reliable, and is suitable for the detection of MP pesticide residues in various vegetables.