Passive optical probes and high-resolution emission spectroscopy are used to provide a general-purpose real-time temperature and chemical species sensing capability. Probes can be inserted in the combustor, at the turbine inlet, in the augmenter, or at the engine exit with application as an engine development diagnostic tool that provides spatially resolved measurements of the key combustion parameters: temperature, CO concentration, and H2O concentration. Multiple probes are arrayed to collect the emitted infrared radiation over different views of the hot gas path. Line-of-sight averaged concentrations and temperatures are determined by spectral analysis of the emitted radiation along each line of sight (LOS). Spatial profiles may also be determined by simultaneous analysis of overlapping lines of sight. The collected infrared spectra contain optically thin and optically thick features that reflect the effects of emission and absorption within the combustion region. The known spectral structure of the component spectra can be used for the automated interpretation of the observed radiance spectra in terms of concentrations and temperatures along the line of sight, and in specific volume elements of overlapping lines of sight. In this work, we present measurements of atmospheric-pressure flames and high-pressure combustors and describe the formalism for fitting the observed spectra to a basis of simulated spectra to extract estimates of concentrations and temperatures. The spectral basis is constructed using a multilayer radiation transport model, in which each line-of-sight or measurement volume is divided into segments of uniform concentration and temperature. The observed radiance emanating from each segment is calculated as a function of the local physical variables. The collection of observed data, which contains a highly structured emission spectrum over each line of sight, is fit to the spectral basis to extract line-of-sight averaged physical properties, or in the case of spatial reconstruction, volume-averaged properties for each of the overlap regions.