Excessive levels of sulfur in the ecosphere are generally found to be detrimental to man and his environment. Four inorganic forms of sulfur are singled out as common constituents of natural and pollutant systems: sulfide (S²⁻), elemental sulfur (S⁰), sulfite (SO₃²⁻), and sulfate (SO₄²⁻). Progress toward characterizing the chemical interactions and toxicity of these species is retarded by the lack of suitable analytical procedures for determining the forms of sulfur in complex solid matrices. An extensive survey of the literature reveals that conventional analyses for sulfur are numerous, but that most techniques are restricted to solutions containing only one type of sulfur. A few complex and cumbersome analytical schemes have been devised, however, for the determination of mixtures of sulfur species in solution. Solubilization of the sample prior to analysis not only adds a step to the procedure, but it also increases the number of variables and the uncertainty associated with the results. The direct determination of sulfur in solids is an alternative to extraction which avoids many of its pitfalls. A critical review of the methods available discloses a paucity of analytical techniques capable of determining individual sulfur species in solid mixtures. X-Ray methods such as X-ray fluorescence and X-ray photoelectron spectroscopy rank the highest as potentially useful probes for eliciting information concerning the oxidation state of sulfur in solids. Routine quantitative work is impeded by the inability to examine bulk properties and the relative lack of sensitivity inherent in X-ray techniques. This apparent analytical void has led to the development of a new technique which can be applied to the determination of inorganic sulfur in solids: time resolved molecular emission spectrometry (TRMES). Solid samples are placed in a small, quartz-lined cavity at the end of a stainless steel rod and the rod is positioned in a low temperature flame, in line with the entrance slit of monochromator. The resulting molecular sulfur emission is monitored at 383.6 nm and is linearly related to sulfur content between at least 10⁻⁵ to 10⁻⁷ g. Qualitative identification of sulfide, elemental sulfur, sulfite, and sulfate is afforded by the separate, time dependent emission responses which are observed. Detection limits are species dependent and theoretically, in a 1.5-mg sample, range from 0.2 pg for sulfite to 8ng for sulfate. In practice, quantitation at these low levels is hindered by inhomogeneous standards. Effects due to different bulk matrices, the presence of two or more sulfur species in a sample, or interferences from counterions are minimal or have not been detected. TRMES provides an alternative to aspirating the sample into the flame and greatly reduces the amount of sample that is required. In addition, the cavity serves to confine the sample and the resultant emission to a small, predetermined region of the flame. This enhances sensitivity and allows one to choose the optimum sample position in the flame. Standard flame emission instruments may be readily adapted to this technique by the addition of a simple, easily fabricated sample introduction device. The utility of TRMES is demonstrated by the determination of sulfur species in several complex matrices, including: coal (S₂²⁻); copper and lead smelter particulates (S²⁻, SO₃²⁻, SO₄²⁻); deep-sea ferromanganese nodules (S⁰); and geologic materials (S²⁻, SO₄²⁻).