Ruthenium catalyzed precursor is the most principal factor affecting the catalytic performance of the supported ruthenium catalyst. Some impurities in the ruthenium catalyzed precursor can inhibit the catalytic performance. In particular, the high content of impurities (such as S, P, Cl and As) reduces activity of the catalyst. In severe cases, the catalyst can be poisoned, thus the level of impurities in the catalytic precursor must be controlled. In this paper, we report an analytical method for rapid and accurate determination of impurity elements in ruthenium nitrosyl nitrate (Ru(NO)(NO3)3) precursor. After dissolved by nitric acid, the impurities (such as P, S, Ti, V, Cr, Mn, Fe and As) were directly determined by inductively coupled plasma-tandem mass spectrometry (ICP-MS/MS). In order to prevent the hydrolysis of Ru(NO)(NO3)3 into Ru(NO)(NO3)x(OH)3-x, we used diluted nitric acid to dissolve the samples while retaining their stability. In the MS/MS mode, the first quadrupole mass filter (Q1) controlled the collision/reaction cell (CRC) ions, only allowing analytes with the same mass charge ratio (m/z) into the CRC. It prevented the interfering ions from the sample matrix and plasma Ar from traveling outside of the CRC, eliminating a significant mass spectral interference. The reaction of target ions P+, S+, Ti+, V+ and As+ with O2 (added into the CRC as a reaction gas) was an exothermic process, which could spontaneously produce corresponding oxides (31P++O2→31P16O++O, ΔHr=-3.17 eV; 32S++O2→32S16O++O, ΔHr=-0.34 eV; 48Ti++O2→48Ti16O++O, ΔHr=-1.63 eV; 51V++O2→51V16O++O, ΔHr=-0.85 eV; 75As++O2→75As16O++O, ΔHr=-0.63 eV). The reaction of Cr+ and Mn+ target ions with O2 was an endothermic process (52Cr++O2→52Cr16O++O, ΔHr=+1.38 eV; 55Mn++O2→55Mn16O++O, ΔHr=+2.15 eV). In order to promote this endothermic reaction, we adjusted parameters of the CRC, in particular, by setting the octopole bias voltage to a negative voltage. Under these conditions, kinetic energy of Cr+ and Mn+ increased and the ions accelerated before the reaction with O2. However, the P+, S+, Ti+, V+, Cr+, Mn+ and As+ ions did not react with O2 in CRC, but still maintained the original m/z. The second quadrupole mass filters (Q2) could block out these interfering ions allowing the oxide-forming ions to enter the detector. This technique eliminated almost all interference from P, S, Ti, V, Cr, Mn and As. NH3 has high reactivity and a pair of lone pairs of electrons, therefore it reacts with many metal ions forming cluster ions. By adding NH3/He as a reactant gas into the CRC, the mass shift reaction of the target Fe+ ions with NH3 occurred. Among the multiple cluster ions, content of Fe(NH3)+2 was the highest and no interference was observed. Thus, we eliminated the interference by the NH3 mass shift method. All 8 elements had a good linear relationship in the range of 0～500 μg·L-1 with the correlation coefficient R2≥0.999 8. The instrumental limit of detection (LOD) of analyte ranged from 0.29 to 485 ng·L-1. According to the established method, the contents of impurity elements in the samples were analyzed. The spiked recovery of the analyte ranged from 93.2% to 107.5%, and the relative standard deviations (RSDs) were less than 3.9%. The proposed method has the advantages of simple sample processing, high speed of analysis and high precision, and is suitable for accurate determination of impurities in Ru(NO)(NO3)3, thereby providing a quality guarantee for preparing the supported ruthenium catalysts.