Abstract: Aromatic parameters have been widely applied to assess the maturity of organic matters. Some parameters may not only be affected by maturation, but also more affected by organic matter source, sedimentary environment, etc. Therefore, the application effect of some parameters in some study areas were not satisfying, and the previous relational expression to calculate the vitrinite reflectance (R_c) which only based on a single parameter is also not suitable for other areas. In this study, we proposed a method to improve the accuracy of using multiple aromatic parameters to quantitatively evaluate source rock maturity. This method was based on a certain number of GC-MS analysis data of aromatic fractions of source rocks in the study area, to find out a most sensitive parameter to maturity in three categories of aromatics (alkyl naphthalene series, alkyl phenanthrene series and alkyl dibenzothiophene series) respectively. Here the most sensitive parameter meant having the largest correlation coefficient to R_o in certain category. Then a multiple linear regression equation was established among these three parameters and R_o using a numerical analysis software, and with this equation R_c can then be calculated with multiple parameters. This method is applied to the study of Mesozoic lacustrine source rocks in the Hari Sag of Yingen-Ejinaqi Basin, the organic matter types of these source rock are mainly type Ⅱ₁-Ⅱ₂. The results indicated that the correlation coefficient (R_c) between R_o and R_c which was calculated by established formula was as high as 0.96, and compared with the single aromatics parameter, the correlation coefficient has been considerably improved.