Colloff et al. in Marine and Freshwater Research (http:dx.doi.org/10.1071/MF14067) examined time-series data for flow-dependent vegetation, invertebrates, fish, frogs, reptiles and waterbirds in the Murray–Darling Basin, 1905–2013. They concluded that temporal patterns fluctuated, declining during droughts and recovering after floods. They suggested that major changes in land use in the late 19th century permanently modified these freshwater ecosystems, irretrievably degrading them before major water diversions. Restoring water to the environment might then be interpreted as not addressing biotic declines. We argue that their conclusions are inadequately supported, although data quality remains patchy and they neglected the influence of hydrology and the timing and extent of water resource development. We are critical of the lack of adequate model specification and the omission of statistical power analyses. We show that declines of native flow-dependent flora and fauna have continued through the 20th and early 21st centuries, in response to multiple factors, including long-term changes in flow regimes. We argue that flow-regime changes have been critical, but not in isolation. So, returning water to the environment is a prerequisite for sustained recovery but governments need to improve monitoring and analyses to adequately determine effectiveness of management of the rivers and wetlands of the Murray–Darling Basin.