This paper describes a modeling study that was undertaken to identify the major ways that coal quality affects alkali vapor emissions in PFBC exhaust systems. The central premise is that the impact of coal quality can be reliably evaluated from kinetically based predictions of the concentrations of NaCl, KC1, SO2. HCI, O2, and H2O in the exhaust from the bubbling bed in combination with ther-mochemical equilibrium relations for conversion of alkali chlorides into alkali sulfate fumes in the freeboard. This hypothesis was implemented in a PFBC simulator that accounts for the major aspects of coal combustion and sulfur capture, and includes several mechanisms for alkali transformations. The model predicts a huge variation in exhaust Na vapor levels for different coals, ranging from the ppm-level with low rank coals to tens-of-ppb with hv bituminous coals for realistic levels of in-bed sulfur capture. This range is consistent with measured Na vapor emissions from two laboratory-scale systems. Hence, this approach exhibits the potential to predict Na-vapor emissions within useful quantitative tolerances at temperatures above 900°C with any coal type that contains sufficient chlorine to shift all alkali vapors into their chloride forms. It also demonstrates that the impact of all the coal properties that affect alkali vapor emissions can be expressed in terms of only two primary variables, the SO2 and HCI concentrations in the exhaust after the alkali sulfate fume has formed. Whereas the essential connections among alkali vapor emissions and the concentrations of alkali vapors. SO2, and HCI are not widely recognized, this study identified a complete set of test conditions that are essential elements for stringently evaluating predicted alkali levels. Simultaneous measurements of alkali vapors, SO2, and HCI are the most important data requirements, together with the Na-, K-, and Cl-levels in the test coals and accurate measurement of the temperature and cooling rate at the sampling point.