The objectives of the MASFILE-I Pilot Project were (1) to test the utility and cost of compiling a manipulative data base from remote card files; (2) to test the utility of the Administrative Terminal System (ATS) for inputting bibliographic data into computer files from catalog card copy at a central location; (3) to test the adequacy of a modified MARC tagging scheme for labelling, inputting, and retrieving formatted bibliographic data elements; (4) to determine overlap of items in the file; (5) to aid the Five Associated University Libraries (F.A.U.L.) in designing a compatible worksheet for transferring the intellectual product of local catalogers into various on-line and off-line machines; (6) to develop recommendations for building a bibliographic data base. A sequential sample of shelf list catalog cards was selected from each F.A. U.L. library in the Library of Congress Classification for the Book Trade and Library Science (Z116-Z1000.5), manually merged, converted to machine readable form by the IBM ATS system at SUNY-Buffalo in a modified MARC-I format. After editing, a list of 1827 items was published containing full citations, holdings statements, and indexes by main entry, LC card number, and LC class number. A draft cataloger's worksheet was designed and is undergoing testing. Overlap studies were made, and time and cost figures compiled. Recommendations for continuation of the project (MASFILE-II) are also included. (Author)