Raman spectroscopy as a rapid，non-destructively testing technique has been widely used in many fields.In this study，49 spectra of gasoline samples were collected with portable Raman spectrometer.Adaptive iteratively reweighted penalized least squares was employed to remove the strong fluorescence background of the spectra.The partial least squares（PLS） calibration model was established in the optimal conditions to predict the content of MTBE.Correlation coefficient，fitting error of training set samples and correlation coefficient，prediction error of test set samples were used to evaluate the quality of the model.The best model showed the satisfactory predictions for the four values mentioned above：0.996 0，0.316 1 and 0.996 6，0.490 1.The results showed that the method was suitable for the fast and reliable determination of the MTBE in gasoline.