A quantitative technology based on near infrared spectroscopy (NIR) is proposed to detect pesticide (chlorpyrifos) residues in vegetables in the paper. There are three groups of samples (totally 100 unit, concentration from 0.1mg/kg to 4mg/kg) used in the experiment. The first group sample is confected to simulate vegetable juice to validate feasibility of the NIR quantitative method. As shown in this work, the first NIR model has good indexes (R2 is 0.9807, RMSEP is 0.149). Next the second group sample is confected by adding material standards of 1 mg/ml of chlorpyrifos to spinage juice. The second NIR model also has good performance (R2 is 0.9858, RMSEP is 0.207). The third group sample is confected as the second group to validate repeatability of the detection method based on NIR. The third model is also excellent (R2 is 0.9986, RMSEP is 0.0485). The exploratory experiment results indicate the detection technology of pesticide residue in vegetables based on NIR is feasible in the given conditions. But supposed to the natural vegetable samples that include pesticide residue, the method should to be validated more.