This study investigated potential impacts of climate change on travel disruption resulting from road closures in two urban watersheds in the Portland metropolitan area. We used ensemble climate change scenarios, a hydrologic model, stream channel survey, a hydraulic model, and a travel forecast model to develop an integrated impact assessment method. High-resolution climate change scenarios are based on the combinations of two emission scenarios and eight general circulation models. The Precipitation-Runoff Modeling System was calibrated and validated for the period 1988-2006, and simulated for determining the probability of floods from 2020-2049. We surveyed stream cross sections at five road crossings for stream channel geometry and determined floodwater surface elevations using the HEC-RAS model. Four of the surveyed bridges and roadways were lower in elevation than the current 100-year floodwater surface elevation, leading to relatively frequent nuisance flooding. These roadway flooding events will become more frequent under some climate change scenarios in the future, but climate change impacts will depend on local geomorphic conditions. While vehicle miles traveled were not significantly affected by road closure, vehicle-hours delay demonstrated a greater impact from road closures, increasing by 10 percent in the Fanno Creek area. Results indicate that any cost analysis is extremely sensitive to the occurrence of human fatalities or injuries and fairly insensitive to delay costs. In addition, this research presents a comprehensive classification of flooding costs, identifies preventative measures, and makes short- and long-term recommendations. Our research demonstrated the usefulness of the integration of top-down and bottom-up approaches in climate change impact assessment, and the need for spatially explicit modeling and participatory planning in flood management and transportation planning under increasing climate uncertainty.