Determinations of low-level Zr, Nb, and Mo in steel samples by glow discharge mass spectrometry are described. The possibilities of metal argide correction on the isotopes of those elements and the use of doubly charged ions are discussed. Selection of suitable isotope(s) depends on the major and minor constituents of the sample. Correction of metal argide interferences MAr+, where M is Fe, Cr, Mn, or Ni, can generally be made by monitoring the ion intensities of the M+ and production rate of FeAr+/Fe+. Use of doubly charged ions for Zr, Nb, and Mo determination is not feasible because of the existence of doubly charged metal argides such as FeAr2+ or low intensities of the doubly charged analyte ions.Recently the requirements for trace analyses for high-purity iron as well as low-alloy steels, which are based on high-purity iron, have been increasing. Traditionally, a variety of analytical methods are applied for the determination of major and trace elements in steels, such as atomic absorption spectrometry (AAS),1 inductively coupled plasma atomic emission spectrometry (ICP-AES),2 X-ray fluorescence spectrometry (XRF),3 or spark emission spectrometry (SES).4 Those methods are well established, and numerous papers were published discussing specific analytical problems. AAS and ICP-AES are good for quantitative analysis with good sensitivities. AAS especially has a superior sensitivity for specific elements, and ICP-AES has the capability of multielement analysis. Both of these methods, however, usually require time-consuming acid-decomposing procedures because of liquid-state sample introduction. It is also difficult to analyze gas components with them. On the other hand, XRF and SES do not require wet chemistry. XRF is good for quantitative analysis, but its sensitivity is somewhat poor. SES is widely used for steel analysis because of its wide-range coverage of the elements and good sensitivity, but it is poor in quantitative analysis.Glow discharge mass spectrometry (GDMS) has been utilized recently for the determination of trace-to ultra-tracelevel impurities in high-purity metals or semiconductors. The reasons for this are as follows: (1) almost all the elements can (1) Japanese Industrial Standard, G1257. Methods for Atomic Absorption Spectrochemical Analysis of Iron and Steel. 1994. ( 2) Japanese Industrial Standard, G1258. Methods for Inductively Coupled Emission Spectrochemical Analysis of Steel. 1989, (3) Japanese Industrial Standard, G1256. Methods for X-ray Fluoresence Spectrometric Analysis of Iron and Steel. 1982. (4) Japanese Industrial Standard, G1253. Methods for Photoelectic Emission Spectrochemical Analysis of Iron and Steel. 1983.