Despite the large body of excellent work that has been done on the quantitative description of char combustion, our ability to specify the conditions to be met in order to maintain the carbon loss from pulverized coal-fired boilers below a desired value is still limited. The problems arise, in part, from the large number of fuel, combustion, and equipment characteristics that may influence carbon burnout. Measurements of the concentration and size distribution of carbon in fly ash from a 30 MW (electric) utility boiler have been used to test assumptions regarding the factors controlling carbon conversion. The principal explanation for unexpectedly high carbon lasses under the conditions investigated was air leakage into the furnace and convection sections, which caused the furnace gas to be richer than expected from the flue gas analysis. Assuming that the inleaking air made no contribution to combustion, a calculation of carbon burnout, using the mean char combustion rate of Hurt and Mitchell [Twenty-Fourth Symposium (International) on Combustion, 1992, pp 1243-1250] and the char reactivity distribution of Hurt, Lunden, Brehob, and Maloney [Twenty-Sixth Symposium (international) on. Combustion, 1996, pp 3169-3177], reproduced the carbon loss and size distribution of unburned particles and was within a factor of 2 of the Babcock & Wilcox correlation of carbon loss with stoichiometric ratio [Steam/Its Generation and Use, 37th ed., 1963, p 17-21]. According to the calculations, the distribution of char reactivities has a significant influence on the carbon loss and abundance of small particles in the unburned carbon size distribution.