As a continuation of research into online public access catalog systems (OPAC's), this study focused on the actual use of OPAC's and on refinement of the methodologies used in carrying out transaction analysis studies. Current patron use patterns were obtained via machine-readable catalogs from four online public catalog systems: The Library of Congress, Syracuse University, Dallas Public Library, and The Ohio State University. Data were also collected in the searching of 76 OCLC public terminals at 53 additional libraries. Such data included terminal identification, user commands and codes, system commands and codes, the search key, session beginning and ending, user and system response times, search text, and dates. Results indicate that accurate information on the amount of time spent at the terminals is unavailable except at The Ohio State University; time spent at terminals is higher than time spent at card catalogs at the same institution; errors frequently occur in sequence; and transaction log analysis is an effective tool for studying user behavior. Utilizing the OPAC transaction tapes, statistical measures have established the relative use of commands within the systems and the sequences of actions to determine user patterns. This report includes 71 illustrations and 46 tables. (Author/LMM)