Climate change may lead to both increased heat and ozone (O 3 ) levels in urban areas over the coming century. To assess potential human health impacts of these changes, models are needed for projecting regional-scale temperature and O 3 changes under climate change, and for characterizing the independent and joint health effects of heat and O 3 . To meet these needs, mortality transfer functions for summer heat and O 3 were developed and applied in a regional health risk assessment for the New York City metropolitan region. The objective was to analyze and project the relative impacts of climate-related changes in mean daily temperature and 1-hour maximum O 3 concentrations on acute non-accidental mortality from all internal causes of death. Exposure-response relationships were developed using a 10-year record of daily summer observations for the region (1990Ð1999). This was done using a time series Poisson regression model that jointly estimated O 3 and temperature effects on mortality, controlling for time trends and day of week effects. To project impacts into future decades, we developed a integrated modeling system that took global scale climate projections for the 2020s, 2050s, and 2080s, using the Intergovernmental Panel on Climate Change (IPCC) A2 and B2 emission scenario assumptions, and down-scaled these to a 36 km grid using regional models for climate and air quality. Regional downscaling was carried out using the GISS-MM5 linked global-regional model system for climate and the Community Multiscale Air Quality (CMAQ) model for air quality. Mortality risks were projected using the transfer functions estimated from the 1990s data. Results showed that both O 3 and heat stress had measurable impacts on mortality risk, but that the relative impacts changed over time. This modeling strategy could be applied in other metropolitan areas and for other health outcomes to assess health impacts of heat and O 3 under a changing climate.