Camellia oil has become one of the most popular edible vegetable oils especially in China. It can be obtained by cold-pressing (CPE), soxhlet extraction (SE), aqueous enzymatic extraction (AEE), and supercritical carbon dioxide extraction (SC-CO2), while research on their efficient identification is limited. Thus, in this study, proton nuclear magnetic resonance (1H NMR) and conventional chemical analysis, respectively coupled to chemometrics, were employed to compare the camellia oils, produced by CPE, SE, AEE, and SC-CO2. The results showed an obviously overlapping among those four different extracted camellia oils, in both principal component analysis (PCA) and hierarchical clustering analysis (HCA), when using fatty acids as input variables. While two obtained PCA models showed good discrimination, according to the minor component compositions (α-tocopherol, squalene, stigmasterol, β-sitosterol, β-amyrin and lanosterol) and 1H NMR spectra, respectively. Additionally, by means of variable importance for the projection (VIP) scores, less 10 dominant 1H NMR spectra signals were screened out as detailed markers for different camellia oils classification. Therefore, 1H NMR combined with chemometrics may be applied as an efficient technique to classify different extracted camellia oils and potentially other vegetable oils.