Raman spectroscopy has been more and more frequently used for pesticide residue detection research in recent years, but the development of sample pretreatment technology is relatively lagging behind. In the study, a rapid method had been developed for the determination of imidacloprid residue in cucumbers with the application of Raman spectroscopy technology and Quick Easy Cheap Effective Rugged and Safe(QuEChERS) sample preparation. Three batches of cucumber samples (the concentration of imidacloprid was within the range of 0.2-5 mg/kg) with different preparation steps(acetonitrile extraction, dehydrate extraction, and fading removing impurity) were chosen as experimental objects. Confocal micro Raman spectrometer was utilized with a 780 nm laser to collect three batches of samples of Raman spectra. Six quantitative prediction models of imidacloprid residue were established based on PLS and PCR methods. The results showed that in addition to the PCR model of the samples by two steps preparation, the residual predictive deviation(RPD) of the other five models was higher than 3. The samples which were only extracted with acetonitrile got the best modeling effect. The correlation coefficient of the calibration set and the prediction set were all above 0.99. The root mean square error of prediction(RMSEP) of PLS method was 0.148 mg/kg, and the RPD was 5.52, which obtained the highest precision of the six predictive models. The results could provide a strong basis for the following-up studies.