Flame chemiluminescence has received increasing attention for its potential sensor and diagnostic applications in combustors. A number of studies have used flame chemiluminescence to monitor flame status, and combustor performance. While most of these studies have been empirical in nature, chemiluminescence modeling has the potential to provide a better understanding of the chemiluminescence processes and their dependence on various combustion operating conditions. The primary objective of this research was to identify and validate the important chemiluminescence reaction mechanisms for OH*, CH* and CO2*. To this end, measurements were performed at various operating conditions, primarily in laminar, premixed flames, fueled with methane, syngas (H2/CO) and Jet-A. The results are compared to 1-d laminar flame simulations employing the chemiluminescence mechanisms. The secondary objective was to use the experiments and validated chemiluminescence reaction mechanisms to evaluate the usefulness of flame chemiluminescence as a combustion diagnostic, particularly for heat release rate and equivalence ratio. The validation studies were able to identify specific mechanisms for OH*, CH* and CO2* that produced excellent agreement with the experimental data in most cases. The mechanisms were able to predict the variation of the chemiluminescence signals with equivalence ratio but not with pressure and reactant preheat. The possible reasons causing this disagreement could be due to the inaccuracies in the basic chemical mechanism used in the simulations, lack of accurate quenching data (for CH*), thermal excitation (for OH*) and radiative trapping (for OH* and CO2*) and interference from the emissions of other species (such as HCO and H2O), for CO2*. Regarding the utility of chemiluminescence for sensing, a number of observations can be made. In syngas-air flames, CO2* is a reasonable heat release rate marker, at least for very lean conditions. OH* shows some advantage in atmospheric-pressure methane and Jet-A flames in general, while CH* is advantageous at high pressure and very lean conditions at atmospheric pressure. The CO2*/OH* intensity ratio is not useful for sensing equivalence ratio in syngas flames, except maybe at very lean conditions. However, the CH*/OH* signal ratio is a promising approach for sensing equivalence ratio at low or very high pressure conditions in hydrocarbon flames. Thermal excitation and self-absorption processes for OH* chemiluminescence can become important for combustors operating at high pressure, high preheat and near stoichiometric conditions. Background subtracted chemiluminescence signals are recommended for sensing purposes.