Water sensitive interventions are being promoted to reduce the adverse impacts of urban development on natural water cycles. However it is currently difficult to know the best strategy for their implementation because current and desired urban water performance is not well quantified. This is particularly at the city-region scale, which is important for strategic urban planning. This work aimed to fill this gap by quantifying the water performance of urban systems within city-regions using ‘urban water metabolism’ evaluation, to inform decisions about water sensitive interventions. To do this we adapted an existing evaluation framework with new methods. In particular, we used land use data for defining system boundaries, and for estimating natural hydrological flows. The criteria for gauging the water performance were water efficiency (in terms of water extracted externally) and hydrological performance (how much natural hydrological flows have changed relative to a nominated pre-urbanised state). We compared these performance criteria for urban systems within three Australian city-regions (South East Queensland, Melbourne and Perth metropolitan areas), under current conditions, and after implementation of example water sensitive interventions (demand management, rainwater/stormwater harvesting, wastewater recycling and increasing perviousness). The respective water efficiencies were found to be 79, 90 and 133 kL/capita/yr. In relation to hydrological performance, stormwater runoff relative to pre-urbanised flows was of most note, estimated to be 2-, 6- and 3- fold, respectively. The estimated performance benefits from water sensitive interventions suggested different priorities for each region, and that combined implementation of a range of interventions may be necessary to make substantive gains in performance. We concluded that the framework is suited to initial screening of the type and scale of water sensitive interventions needed to achieve desired water performance objectives.