1. Reinstating more natural water regimes is often a priority intervention to rehabilitate wetlands that have been degraded through anthropogenic changes to their natural wetting and drying cycles. Hydrological interventions are often made in chronically desiccated wetlands but less commonly in wetlands that have been permanently inundated and that require a drawdown phase for rehabilitation. Reports on the effectiveness of reinstating a drawdown phase in chronically inundated wetlands are particularly rare. 2. We undertook a landscape-scale, experimental drawdown of water levels at Dowd Morass, a large, Ramsar-listed, brackish-water wetland in south-eastern Australia that had been artificially flooded for 30+ years. During the hydrological manipulation, c. 500 ha of the wetland was drawn down and re-flooded, and the remaining c. 1000 ha was used as a control site. Fringing areas with a fluctuating water regime were used as a reference site. Results were analysed in terms of gradient analysis, by classifying the different water regimes created by the hydrological interventions. The response of wetland vegetation was measured along replicated transects over a 4-year period, before, during and after drawdown. Wetland plants were assigned to plant functional groups for analysis. Assembly theory and knowledge of life-history traits were used to predict that drawdown would promote recruitment of plant species that required exposed sediment for germination and seedling establishment. 3. Within-wetland microtopography interacted with the hydrological interventions to generate three distinct water regimes, which were differentiated by the spatial extent of exposed sediment and duration of the dry period. Drawdown promoted limited recruitment of some plant species, and the survival of cohorts then depended strongly on the extent and duration of the dry period. Species richness and vegetation cover (understorey and overstorey) continued to decline in constantly flooded areas of the wetland. Increased salinisation of sediments and surface waters reduced the effectiveness of the drawdown and dramatically affected species richness and cover of aquatic vegetation, which did not recover fully when fresher conditions returned. 4. The capacity of vegetation to respond to the reinstatement of a drawdown cycle following chronic inundation was constrained by abiotic (e. g. salinity) and biotic (e. g. depauperate seedbanks) factors. Reinstating a dry phase in chronically inundated, brackish-water wetlands is complex and risky and may not effectively improve vegetation condition in the short term. In the case of Dowd Morass, rehabilitation was most successful in sites that had been shallowly flooded prior to drawdown and that remained dry for longest.