Internal standard lines play a crucial role in the univariate quantitative analysis in portable laser-induced breakdown spectroscopy (LIBS) technology. To overcome the uncertainty of the conventional internal standard method, the distribution principles of high-quality internal standard lines were studied and revealed at the macro and micro levels, and an automatic internal standard line selection method based on the Q-value was proposed. Using this method, in the quantitative analysis of Si in low-alloy steel samples, the average relative error of cross-validation (ARECV), root mean squared error of cross-validation (RMSECV), and the limit of detection (LoD) were decreased significantly from 27.42%, 0.041 wt%, and 1060 μg g-1 to 18.65%, 0.026 wt%, and 680 μg g-1, respectively. The quantitative analysis results of Cr, Cu and Ni showed that it has excellent generalization ability. The results indicate that this method can screen out the optimal internal standard lines efficiently and accurately, which provides a new approach to improve the performance of univariate quantitative analysis in portable LIBS.