The North American Regional Climate Change Assessment Program (NARCCAP) is constructing projections of regional climate change over the coterminous United States and Canada in order to provide climate change information at decision relevant scales.  An important goal of NARCCAP is to establish uncertainties in regional scale projections of future climate by using multiple regional climate models (RCMs) nested within multiple atmosphere-ocean general circulation models (AOGCMs).  NARCCAP is using six nested regional climate models at 50 km resolution to dynamically downscale realizations of current climate (1971-2000) and future climate (2041-2070) from four AOGCMs.  Global time slice simulations, also at 50 km resolution, also are performed using the GFDL AM2.1 and NCAR CAM3.1 atmospheric models forced by the AOGCM surface temperatures and will be compared with with the regional models.  Results from this multiple-RCM, multiple-AOGCM suite will be analyzed to investigate the cascade of uncertainty as one type of model draws information from another.  All output is being made available to the broader community through an archiving and data distribution plan.  NARCCAP will furnish the climate impacts community with data at unprecedented spatial and temporal (hourly to six-hourly) resolution in order provide decision-relevant information for public policy. Simulations also are conducted that nest the participating RCMs within reanalyses of observations.  This can be viewed as nesting the RCMs within a GCM that is nearly perfect (constrained by available observations), allowing us to separate errors attributable to the RCMs from those attributable to the driving AOGCMs.  Results indicate that skill is greater in winter than in summer, and greater for temperature than for precipitation.  Temperature and precipitation errors are uncorrelated from model to model; consequently, the multi-model ensemble has more robust skill than any single model.