The fate of injectant coal in blast furnaces has been investigated through characterizing samples collected from commercial-scale blast furnaces and laboratory-scale tests. Representative samples, including an injectant coal and the corresponding carryover dusts from a full-sized blast furnace, and a set of tuyere level core-drilled samples from another full-sized blast furnace, have been examined. The samples were extracted using 1-methyl-2-pyrrolidinone (NMP) as the solvent, and the extracts were then studied using size-exclusion chromatography (SEC). The carryover dust extracts were found to contain heavy carbonaceous materials of apparent mass similar to 10(7)-10(8) units, on the basis of polystyrene calibration. Similar materials were also found to be present in the NMP extracts of the core-drilled samples taken from the bosh, rear of bird's nest, and deadman regions of the furnace, at the tuyere level. In contrast, the feed coke did not give any extractable material. Controlled pyrolysis and combustion experiments, in an electrically heated wire-mesh reactor, suggest that the extent of injectant coal combustion in the raceway is limited by a very short exposure time to high temperatures and poor oxygen availability. These observations suggest that some coal char particles might escape from the raceway incompletely pyrolyzed. Unburned volatiles, particularly tars, may be further thermally altered, giving rise to the formation of high-molecular-weight (soot-like) materials.