AIRS/AMSU is a state of the art high spectral resolution IR/MW sounding system launched on EOS Aqua in May 2002. The primary purposes of the AIRS/AMSU are for use in improving numerical weather prediction and the study of climate variability and trends. The Goddard DAAC has been analyzing AIRS/AMSU data using the AIRS Science Team Version 4 retrieval algorithm since April 2005. All AIRS/AMSU data will be reprocessed at the Goddard DAAC using the AIRS Science Team Version 5 algorithm starting in late 2006. The AIRS Science Team Version 5 retrieval algorithm contains further improvement in retrieval methodology, including accounting for effects of non-LTE on the channel radiances and improved treatment of surface emissivity effects. In addition, considerable improvement has been made with regard to error estimates and quality control of both the derived soundings and the clear column radiances used to generate them. Improved error estimates and quality control are significant for both data assimilation and climate purposes. The improvements in sounding methodology will be described. Results will be shown demonstrating the accuracy of soundings and clear column radiances and their error estimates, as well as the characteristics of the new quality control algorithm. Results of forecast impact studies using this data will also be presented, assimilating global quality controlled AIRS products with the NASA GEOS-5 Data Assimilation System. Significant positive forecast impact is found in both the Northern and Southern Hemisphere Extra-tropics.