The objective of the project was to develop an inexpensive and effective technique for matching records in multiple fields in order to build a union catalog containing only one representation of a given book. The Information Retrieval Research Lab (IRRL) developed a prototype system called IUCS (IRRL Union Catalog System) by testing various techniques for matching records against large test files of machine-readable records from major institutions that maintain their catalogs in machine-readable form. IUCS involves four major steps: (1) a pre-processing step which converts incoming files into IUCS form; (2) a first pass screen which matches records according to an efficient title-date matching key; (3) a second pass screen which provides more refined matching of the potential duplicates located in the first pass; and (4) a post-processing step which merges the new records into the existing MASTER file to create a union catalog. These procedures were tested by generating a union catalog of three sample files from Ohio College Library Center, Northwestern University, and the University of Chicago. The feasibility study proved it is possible to automatically generate a union catalog with few mismatches or missed matches, and to create and update this catalog economically. (Author/JAB)