Abstract: As an alternative fuel to gasoline, methanol gasoline is economical and environmentally friendly, and its methanol concentration has a significant effect on calorific value, explosion resistance, and combustion efficiency. As a nondestructive, rapid, portable, easy-to-operate, and low-cost analytical method, portable Raman spectroscopy has been applied to the analysis of the methanol content of methanol gasoline. However, the nonlinear Raman frequency-shift demonstration error of portable Raman spectrometers affects the accuracy of analytical modeling. A method for methanol content analysis of methanol gasoline with portable Raman spectrometer calibration is proposed. The polystyrene Raman frequency-shifted reference material is used for multi-point calibration, and a peak detection algorithm is used for the calibration of Raman spectra. Based on two characteristic peaks of methanol, a binary linear regression method is proposed to establish a quantitative analysis model. The prediction results and evaluation indexes of the model show that this method can predict the methanol content more accurately.