Efficiencies of He/NH3 and He/H2 collision gases were compared in a conventional type of hexapole cell of an inductively coupled plasma mass spectrometer (ICP-MS). The optimum conditions [hexapole and quadrupole bias voltage (V(H) and V(Q)) and collision/reaction gas flow rates] were tested for vanadium determination (51V) in chloride matrices. When the He/H2 mixture was used, the optimum values of V(H) and V(Q) were -10.0 and -8.0 V, respectively. This set-up corresponds to the kinetic energy discrimination effect. When the He/NH3 mixture was used, the optimum values of V(H) and V(Q) were +10.0 and -7.0 V, respectively. Positive V(H) values correspond to the ion kinetic energy effect, which allows the reactivity of the ions entering the collision/reaction cell with the reaction gas to be controlled. The obtained results showed that the He/H2 mixture is not optimal for V determination in samples containing chlorides due to the insufficient suppression of the polyatomic interference of 35Cl16O+. Data obtained from vanadium determination using the He/NH3 mixture were consistent for all selected Cl- concentrations, and the results were acceptable. The detection limit was comparable with detection limits obtained from ICP-MS equipped with a dynamic reaction cell. Analyses of elements forming interfering molecules, e.g., iron (56Fe), arsenic (75As) and selenium (80Se), were in good agreement with the certified values for both studied collision/reaction gas mixtures.