Combinatorial libraries of both nucleic acids and peptides have been widely used in many areas of biomedical research and have played a vital role in their progress during the past two decades. Important discoveries have been made using combinatorial libraries, including the identification of many DNA-binding proteins, catalytic RNAs, antibody epitopes, and protein ligands. Most combinatorial libraries used to date are based on chemically synthesized """"""""short"""""""" peptides or oligonucleotides, which are relatively simple compared with the complexities of biologically active molecules. We have devised a novel enzymatic approach of generating complex combinatorial libraries of polynucleotides of continuous open reading frame that can encode long polypeptides. This strategy is highly versatile and powerful, yet simple. Combinatorial libraries of randomly generated complex polypeptides have broad potential applications. In this study, we propose to further characterize this novel strategy to better understand its full potentials and limitations. The scope of this exploratory proposal is limited to the following three specific aims: (1) to characterize and refine the process of generating random polynucleotides; (2) to determine feasibility of controlling biochemical properties of randomly generated polypeptides; and (3) to develop an in vitro transcription and translation system for selecting random polypeptides.