The superensemble methodology incorporates results from a suite of multi models to provide forecasts for hurricane track and intensity. This includes a training phase and a forecast phase. This training phase is based on nearly 100 past forecasts for each of the member models. The training phase generates statistical coefficients that provide collective bias corrections for each of the member models. These statistics are then passed on to the forecast phase where the member model forecasts are corrected to construct a superensemble forecast. The new aspect of this study is the inclusion of a suite of mesoscale models, which appeared necessary for the intensity forecasts. The intensity forecasts based on larger scale models are known to carry very low skills for intensity forecasting. The present study includes the WRF-ARW, ETA, GFDL, MM5, and the official National Hurricane Center forecast (OFCL). The resolutions of these models vary between .08 degrees to .41 degrees. The training phase of this study includes 25 storms from the years 2002 to 2006, adding up to a total of 92 forecasts per member model. Two day forecasts were carried out during the training and forecast phases. This presentation will include the performance of the mesoscale models, the ensemble mean, and the superensemble. This performance is assessed for skill metrics with absolute errors for the tracks and intensity forecasts.