The aims of this project were to: 1. further develop the evaluation systems of Westman (1986), Chapman and Underwood (2000) and Wilkins et al (2004) proposed for the assessment of restored ecosystems; and 2. use these developments to evaluate whether the revegetation of agricultural land on the Cumberland Plain, west of Sydney, NSW, has led to the re-establishment of a grassy woodland. The evaluation system developed in this Thesis was designed to compare three key ecosystem attributes. First, to assess how restoration was progressing, the species richness, composition and vegetation structure of abandoned pasture (starting point), was compared to that of restored vegetation of differing ages (putative mid points), and remnants (goal or end point). Refinements of the previous assessment models included formulation of predictions about native and exotic species richness and composition under the assumption that restoration was succeeding, and explicit testing of these predictions by planned comparisons and trajectory analysis of species composition. Second, the small-scale effects of planted tree canopies on species composition were assessed to test the hypothesis that native tree canopies facilitate the return of natives. Third, the effects of fire and neighbour removal on seedling emergence and establishment in pasture, restored vegetation and remnants were examined to explore what factors controlled germination and establishment. The results of this study indicate that to date, there has been a partial success of the restoration program at the study sites: while native species have returned unaided to restored sites, the trajectory of native species composition was not in the direction of remnants. There was however, increased species richness of exotic species detected underneath planted tree canopies. Patterns of seedling emergence observed in this study suggest that recruitment plays a role in the maintenance of the species composition found in restored vegetation, with seedling emergence dominated by exotics The evaluation methodology developed within this Thesis is a transparent and accurate way to measure ecological changes in vegetation that have occurred as a result of restoration.The restoration evaluation methodology further developed here will be useful to an industry that involves tree planting, landcare, revegetation and bush regeneration. It will complement guidelines provided by government and other sources that advise on practical aspects of revegetation and will be one of the few which have examined the success of revegetation in ecological terms that are founded on sound scientific basis. / Doctor of Philosophy (PhD)