Plasma in contact with liquid is widely studied in research involving water purification, nanoparticle synthesis, biomedical applications, and so on. However, plasma–liquid interactions, which are physical and chemical phenomena occurring at a plasma–liquid interface, have not yet been fully understood. To deepen our understanding of the interactions, in this study, we have investigated where specifically short-lived active species generated by plasma react with solutes at the plasma–liquid interface. We adopted chloride (Cl − ), bromide (Br − ) and iodide (I − ) ions as indicators of the reaction region of the short-lived active species, because it is known that these halide ions are located closer to the gas–liquid interface in an order of I − , Br − , and Cl − . A DC argon plasma was irradiated to 2.1 mol l −1 sodium chloride (NaCl), bromide (NaBr), and iodide (NaI) solutions, and we found that chlorine was not detected after 600 s plasma irradiation while bromine and iodine were observed after 30 s plasma irradiation. To explain the different characteristics of halogen (chlorine, bromine, and iodine) generation, we postulated that, because of the surface activity and low kinetic energy of the short-lived active species generated by plasma, they could react with the solutes only at the topmost layer of the water surface, which is an atomic monolayer in liquid in contact with the gas phase. Based on our suggestion, we could say that the Cl − is unable to react with the active species into the chlorine because the ions were excluded from the topmost layer, and that, by contrast, the Br − and I − were able to react into the bromine and iodine, respectively, because some of these ions were located in the topmost layer. For understanding the plasma–liquid interactions more, a difference between chemical reactions induced by plasma and conventional electrolysis has been also discussed.