Presented at the 2014 Aquarius/SAC-D Science Team MeetingThe CONAE MWR team has developed an improved algorithm for retrieving ocean wind speed using the newest MWR V7.0 brightness temperature data. This poster presents a description of the algorithm and results of a comprehensive on-orbit validation using coincident ocean wind speed retrievals provided by Remote Sensing Systems, Inc (RemSS).The previous algorithm was based on MWR ocean Tb measurements @ 37 GHz (52° and 58° incidence angles), ancillary Aquarius (GDAS) SST and the microwave radiative transfer theory developed by Wentz, 1992. A linear regression was used to translate the MWR Tb @ 37 GHz V-&H-pol to match WindSat brightness temperature at 53° incidence angle. During the on-orbit Cal/Val period, a simple regression was applied to remove the mean wind speed biases compared to collocated WindSat wind speed retrievals.The new model is also based on the same microwave radiative transfer theory, but now all coefficients have been tuned using one year of MWR brightness temperature at 23 H (H-pol) and 36.5 GHz (H- & V-pol) and auxiliary data, such as sea surface temperature obtained from GDAS (AQ L-2 product) and wind direction from NCEP. Unlike the previous algorithm, this version makes a wind direction correction to the surface Tb's before retrieving atmospheric transmissivity and the isotropic ocean surface wind speed at 10 m height, and has two different sets of coeffecients for each incidence angle.For on-orbit validation, results for selected MWR single pass comparisons with independent collocated ocean wind speed retrievals provided by RemSS are presented. Also presented are statistical results of global MWR wind speed averages in different seasons. Results demonstrate significant improvements in the random differences between MWR retreivals and the RemSS surface truth.