MARC II database statistics and trends from the years 1973-74 to 1980-81 are presented for the benefit of processors of the MARC file who may use the data for planning file structures, selecting subsets of MARC for local processing, estimating processing time based on record sizes, and estimating future file sizes based on growth rates. It is noted that the statistics can also be valuable in the preparation of user profiles and search strategies for subject searching of the MARC database online, and in the establishment of content definitions for full and abridged bibliographic records. Tables included in the report provide a detailed characterization of the content and changes in the MARC database in terms of such basic attributes as record length, field tag occurrence, data element length per field tag, and distribution of records by Dewey Decimal Division and Library of Congress class code. Distributions for Library of Congress categories vs. mean record length, and main entry length vs. mean record length, are also given. A brief analysis of trends precedes the report's extensive tables. (Author/ESR)