Lithium is an important strategic metal and new energy material, and its development and utilization has attracted extensive attention worldwide. There are huge amounts of lithium resources in high salinity brines especially in salt lake brines and underground brines. It is necessary to accurately determine lithium content during the development of these resources. However, high concentrations of Na+, K+, Ca2+ and Mg2+ in brines will result in serious interference with the determination of trace lithium. Inductively coupled plasma optical emission spectrometry (ICP-OES) has the ability of multi-element simultaneous analysis with wide linear range. In order to realize swift and accurate determination of lithium in brines, the analysis of lithium in high salinity samples by ICP-OES was carried out in detail in this work. Results showed that lithium has a higher signal to noise ratio at 610.364 nm, where no obvious spectral line interferences of Na+, K+, Ca2+, Mg2+ and Ar were presented for the determination of lithium. However, large amount of Na+, K+ and Mg2+ coexisting in the samples would lead to positive matrix interferences, while the Ca2+ caused negative interference. The internal standard method widely used to eliminate matrix interferences could not effectively solve this problem using whether yttrium or scandium as an internal standard element. Because the standard addition method involves complex operations and is not suitable for batch sample analysis, while the matrix matching method must match these interfering components with corresponding ions and thus is not suitable for the analysis of batch samples with different matrix compositions, the feasibility to match complex matrix with a single component was investigated in this work. Since NaCl widely exists in brines and has obviously sensitizing effect on lithium determination, we found after a series of studies that when the total amounts of NaCl, KCl, MgCl2 in samples were no more than 40 g·L-1, the interferences of these coexisting ions could be successfully solved by adding 10 g·L-1 NaCl into both the samples and the standard solutions. Although the negative interference of Ca2+ could not be effectively solved by this method or pre-separation by precipitation, it did not lead to obvious impact on the determination when its concentration was no more than 1.8 g·L-1. When this method was used for the determination of three kinds of spiked samples with different matrices, the recoveries were within the range of 96.60%～104.20%. Meanwhile, the accuracy and reliability of this developed method were also demonstrated by the comparison with these obtained by the inductively coupled plasma mass spectrometry (ICP-MS) method (The relative errors varied within ±3.66%). Only single NaCl was used in this proposed method to match different coexisting ions, by which it not only simplified the operation, but also realized the batch analysis of samples with different matrix compositions. Therefore, it has important significance for the swift and accurate determination of lithium in brines as well as for the development and utilization of lithium resources in brines.