The purpose of the Library of Congress Machine-Readable Cataloging (MARC) pilot project was to provide libraries with catalog data encoded on magnetic tape in order to determine the feasibility of centralized computerized cataloging in support of library functions. Sixteen participants were selected on the basis of expressed interest, available staff, access to computers and associated equipment, and proposed applications. The weekly tapes were used by the participants for two basic purposes: (1) using the data in everyday library processing, and (2) using the data to provide extra services such as machine searching. Experiences at the University of California, Los Angeles (UCLA), which used the MARC information to produce catalog cards and alphabetize subject lists for use in book selection, indicated that there were problem areas both in using the programs provided and in creating new programs for processing tapes organized in the MARC format. In general the experiments at UCLA demonstrated that (1) it takes longer to plan and implement automated library procedures than most library administrators suspect, and (2) a full-scale nationwide MARC distribution service could not be realistically utilized without a substantial "lead time" for libraries to prepare themselves for processing the data. A listing of the programs developed at UCLA is also given. (CM)