Aiming at the problem of making and selling fake peony seed oil in the market, in situ, Raman spectroscopy and chemometrics were used to rapidly classify and identify the authenticity of peony seed oil. Peony seed oil mixtures with different adulteration concentrations were prepared, and their Raman spectra were collected. The spectral data were de-baseline processed by a wavelet algorithm. The characteristic information of the Raman spectrum was extracted based on principal component analysis. The adulteration quantitative models of peony seed oil and rapeseed oil were established by using three chemometrics methods, including multiple linear regression (MLR), principal component regression (PCR), and partial least squares regression (PLSR). The R² of MLR, PCR, and PLSR were 0.9675, 0.9839, and 0.9846, respectively, and the RMSE were 0.057, 0.041, and 0.040, respectively. The results show that the PLSR algorithm has the best prediction effect. In this paper, a rapid detection method for the authenticity and adulteration of peony seed oil using in-situ Raman spectroscopy and chemometrics was proposed and implemented, which provides methodological reference and guidance for the rapid detection of peony seed oil adulteration.