Fish responses to flow regimes can be used to assess potential vulnerability of stream ecosystems to future flow alteration. Importantly, ecological flow relationships are hypothesized to vary regionally and our understanding the benefits from landscape-scale research that addresses complex variation in climatological, physical, and biological components. The U.S. Geological Survey (USGS) has identified priority ecosystems as regions that can be used for landscape-scale climate change vulnerability assessments. Ongoing work demonstrates how these USGS large landscapes can be used to facilitate partner-informed ecological flow vulnerability assessments and serve as potential pilot areas for larger scale ecological flow modeling and monitoring efforts. We highlight a collaborative research project bringing together ecologists, geospatial scientists, hydrologists, and partners working across scales in five USGS large landscapes: the Columbia, Upper Colorado River, and Mobile Basins, and the Great Lakes and Chesapeake Bay watersheds. Outreach and engagement efforts with local to national partners (government, non-governmental, and tribal) identified science needs and geospatial scientists summarized contemporary and future landscape and environmental datasets for hydrology-informed machine-learning models that will predict time-varying streamflow characteristics (the magnitude, frequency, duration, and timing of droughts and floods) at ungaged locations and project them to future conditions. Ecologists will use estimated streamflow characteristics to perform vulnerability assessments of lotic fish to future flow alteration, as informed by partner needs and feedback. We’ll discuss opportunities and challenges in designing landscape-scale vulnerability assessments to meet science needs from local to national scales.