To develop a method for quantifying the vegetation of Pennsylvania-age coal beds, of four coal-ball (permineralized peat) profiles and four coal column samples from the Herrin coal bed (Kentucky No. 11) Carbondale Formation in western Kentucky were compared. An estimated 89.5% of the coal can be identified botanically. Compaction ratios for individual tissues were estimated using point counts of organic matter in coal balls. The estimated abundances of major plant groups (lycopods, ferns, sphenopsids, and pteridosperms) in coal balls differ by less than 10% compared to coal after accounting for differential compaction of plant tissues. Standard deviations in taxonomic and maceral composition among coal columns are generally less than 2%. Consistent differences in botanical composition were found between benches showing that the method is consistent when applied to sufficient thicknesses of coal. It was not possible to make fine-scale correlations within the coal bed using the vegetational data; either the flora varied considerably from place to place or the method of quantification is unreliable for small increments of coal (5 cm or less). In the coal, pteridosperm abundance is positively correlated with underlying shale partings. This correlation suggests that pteridosperms are favored either by higher nutrient levels or disturbance. In the third of four benches in the Herrin coal bed, a succession from Sigillaria-containing zones to zones dominated by Lepidophloios hallii is interpreted as a shift towards wetter conditions. In the other benches, the main factors controlling the taxonomic composition appear to have been the relative abundance of nutrients and/or the frequency of disturbance as indicated by the relative abundance of partings. Criteria for distinguishing between domed and planar swamps are discussed. These include: distribution of partings, type of plant succession, and changes in plant diversity, average plant size, preservational quality and sporinite content. The infrequency of partings in bench C suggests a peat dome developed while the peat of that bench was accumulating but other evidence either fails to support the development of a peat dome or is ambiguous. The maceral composition resembles those of other Carboniferous coals which are thought to have formed from planar peat swamps. Formation of fusain bands appears to be associated with processes occurring above the peat surface, such as burning or prolonged oxidative exposure. Oxidation of accumulated peat is unlikely because fusain bands rarely include more than a single plant. ?? 1988.