Glow discharge spectrometry (GDS) is a technique for rapid depth profiling and bulk analysis of materials. The latest instruments can be operated in either d.c. or r.f. modes. The d.c. mode provides analysis of conductive materials only, while the r.f. mode allows analysis of both conductive and non-conductive samples. Until now, no theory for r.f.-GDS has been available. The recently proposed d.c. theory, which assumes that emission yield depends only on pressure, is extended to r.f.-GDS and is shown to be valid for the r.f. mode as well. A key to tbe success of the theory is that, in support of Duckworth and Marcus, the two modes are shown to have similar analytical characteristics, e.g. in comparing the two modes most elements examined had similar sensitivity factors and background equivalent concentrations. This will facilitate calibration of the light elements hydrogen, oxygen and nitrogen for use in the d.c. mode. Recommended procedures for bulk analysis, in either d.c. or rf. modes, are presented and illustrated for manganese in steel. Both modes provide calibration curves of high linearity. Quantitative depth profiles of three commercial metallic-coated steel samples are compared in d.c. and r.f. modes and the similarity of the quantitative profiles, given tbe very different lamp parameters operating between the modes, confirms that the theory works for both d.c. and r.f. The theory is then applied to data from a commercial prepainted metallic-coated sample to provide the first quantitative GDS depth profile of a non-conductive material.