This thesis examines the effects of species, rainfall and soil type on tree biomass regressions, as well as the effects of stand dominance and structure on stand biomass regressions in north-east Australian woodlands. This was achieved by examining tree characteristics and biomass relationships for a series of woodland monitoring sites throughout the study area. This study utilised a modified data set from this permanent monitoring site network to provide structural attributes for trees and communities of varying composition in the grazed woodlands. These data were supplemented with environmental data and tree harvest data sets. Initially, the research reported in this thesis developed allometric and stand biomass regressions for Callitris glaucophylla communities. This research also demonstrated that changes in tree-form were not reflected in changes in the environment, nor did such changes reflect changes in tree biomass regressions for three eucalypt species. As a result, a common regression provides a robust estimate of total aboveground biomass of eucalypt trees in the study area. Thus expensive destructive harvesting can generally be avoided for minor eucalypt species. Finally, this study demonstrated a successful methodology that described the stand structure of all the grazed woodland sites based on tree heights. This methodology was developed to allow the expansion of a single stand regression to estimate stand biomass across the entire north-east Australian woodlands. The findings demonstrated in this study, combined with the long-term data from the permanent monitoring network sites, should enhance the estimation of carbon flux within eucalypt communities of north-east Australia’s grazed woodlands.