Taphonomic processes that may bias plant-fossil assemblages are important in explaining many aspects of the Cenozoic plant fossil record. Two tree genera, Eucalyptus and Nothofagus, are especially important for understanding the evolution of Australia's forests, because the former currently dominates the continent, whereas the latter is dominant in Cenozoic microfloras. Nothofagus also is prominent in temperate forests of South America and New Zealand, and in the Cenozoic history of these land-masses and Antarctica. This study measures standing biomass and leaf-litter production in contiguous cool temperate rainforest dominated by nanophyllous Nothofagus cunninghamii and wet sclerophyll forest dominated by macrophyllous Eucalyptus regnans in southeastern Australia. It tests an a priori hypothesis that the under-representation of Eucalyptus in the Australian Cenozoic record may reflect differential organ production by Eucalyptus and Nothofagus, respectively. A taphonomic bias was observed between the standing crop and leaf-litter production among the five canopy species that were examined. Nothofagus cunninghamii has particularly high levels of leaf production relative to standing biomass, while E. regnans has low levels of leaf production relative to its standing biomass. It also was found that when there are large differences in leaf size, leaf counts do not accurately reflect rank-order dominance patterns in the standing vegetation, whereas measurements of leaf area do reflect the general dominance patterns. These data are used to reevaluate the earliest reliable Australian Eucalyptus fossil locality, the Late Oligocene Berwick Quarry Flora, which contained both Nothofagus leaves and pollen. This reevaluation indicates that Nothofagus was present in, but did not dominate, the original vegetation, and Eucalyptus may have been more common than previously considered.