Wetland vegetation communities are mostly composed of species that depend on moist conditions or flooding for part, or all their life cycle. Their structure and species composition have been shown to change in response to hydrologic alterations, with increased hydrological connectivity and flooding driving floristic structure and composition towards more flood-dependent species, while increased drying promotes the establishment of more terrestrial species. Many human-related alterations to the environment act to degrade wetland ecosystems causing shifts in the structure and species composition in plant communities that can be quantified. Such shifts can be used as an indicator to evaluate best management practices for wetland resources such as environmental water, and to assess the efficacy of restoration actions and mitigation activities. We quantified a range of condition-indicator classes for wetland vegetation communities based on floristic composition and structure, in two large inland floodplain wetlands of NSW. We modelled the relationship between condition class and hydrologic and abiotic variables to build ecological reference models for each vegetation community. By quantifying reference conditions in relationship to ecological drivers, as well as conditions at degraded states that deviate from reference conditions, our results show that robust quantitative ecological reference models provide an advantage in temporally changing wetland vegetation communities in contrast to other approaches that use reference conditions. Our ERMs can provide tools for long-term, cost-effective management of environmental water in these wetlands.