This is a proposal to establish the technological proof-of-principle for a high-throughput and parallizable means for generating fully human monoclonal antibodies. We describe the development of a new version of the yeast two-hybrid (Y2H) system to identify antibodies that interact with a bait protein, a novel iterative method for increased searching of sequence space, and a method with which to quickly, cheaply and accurately identify specific antibody sequences giving rise to antibody-antigen interactions. The antibodies we derive are designed to be cloned into frameworks for the immediate development of clinically useful antibody drugs. ? ?