Questions: How does ground-layer plant composition respond to the imposition of woodland habitat restoration treatments following the removal of long-term pastoral management? Do different vegetation types have different trajectories of change? Location: A long-term ecological research site comprising temperate eucalypt grassy woodland and forest in south-eastern Australia, converted from pastoral use to conservation management. This has involved the implementation of habitat restoration treatments in addition to the cessation of livestock grazing and fertilizer inputs. Methods: We surveyed ground-layer floristic composition over a 4-yr interval, in 96 1-ha sites that had been subjected to combinations of restoration treatments aimed at improving the condition of the woodland habitat for fauna: (1) Reduced kangaroo grazing intensity; (2) Addition of coarse woody debris; and (3) Burning. We used nonmetric multidimensional scaling (NMDS) to plot the sites, and to analyse the average trajectories of floristic change for each of six initially identified vegetation types, and the restoration treatments. We partitioned Bray-Curtis dissimilarity measures to rank individual species' contributions to floristic change in each vegetation type between 2007 and 2011. Results: The identities of the six vegetation types identified at time zero was strongly retained over the 4-yr observation period. There was generally a uniform change across the vegetation types, which was the result of increases in biomass of the characteristic dominant perennial grasses, and increases in some native forbs. The two vegetation types with evidence of past fertilization showed some convergence with native vegetation types due to the decrease in some exotic species. The restoration treatments had no significant influence on species composition. Conclusions: The native herbaceous vegetation in the woodland was largely resistant to compositional change, although seasonal conditions increased the biomass of the perennial grasses defining the different vegetation types. However, sites with an evident history of past fertilization and notable amounts of exotic pasture species changed slightly but significantly, with some pasture species declining and floristic convergence with native vegetation over 4 yr. These rates of change need to be considered in the application of vegetation offsetting policy.