1. The mechanisms of community assembly underpinning plant invasion are contested; limiting similarity predicts divergence in the traits of native and alien species whilst habitat filtering predicts trait convergence. Using site‐based floristic data collected three decades apart, we ask whether the competitor–stress tolerator–ruderal (CSR) strategies of herbaceous alien and native plant species in Australian woodlands show evidence of divergence (indicating limiting similarity) or convergence (indicating habitat filtering). We also ask whether increasing dereliction between the two sampling periods, as a result of declining fire frequency, has influenced habitat filtering processes.2. Null models were used to assess the similarity in CSR strategies of native and alien species in order to identify the community assembly processes driving woodland invasions. The observed frequencies of CSR types were also compared to randomly assembled communities from null models for species experiencing increases, decreases or static abundance changes between the two surveys to assess the impacts of dereliction on habitat filters.3. The spectrum of CSR strategies of the most common aliens (n = 25) was highly convergent with the strategies of the most common natives (n = 51). Competitive–ruderal species dominated the flora (> 30% of all species), and strict competitors were absent. No aliens exhibited a stress‐tolerant strategy, and ruderals were more common in the alien flora (i.e. 36% vs. 18% of native species). Null models revealed differences in the CSR strategies of increaser and decreaser species at three of the ten sites examined, although none of these trends were consistent across the sites, indicating that increasing dereliction has not dramatically altered the habitat filters determining species admission.4. Synthesis. The similarity of CSR types between alien and native species indicates that habitat filtering is the principal community assembly process operating in the study region. We discuss our findings with respect to suggestions that limiting similarity and environmental filtering may be occurring simultaneously to structure plant communities. Species abundance changes due to dereliction were not evident as a coordinated shift in the functional composition of woodland dominants suggesting that time lags between altered management regimes (dereliction) and the functional response of species to habitat filters may still be playing out.