MARC 2 1 is a combination of two similar formats, Canadian Machine-readable Cataloguing Format (CAN/MARC) and American Machine-readable Cataloguing Format (USMARC), and then the differences are eliminated. Based on the international standard Z39.2 of American National Standards Institute -ANSI, it integrates MARC formats of various countries, breaks the boundaries of various countries by using the Internet and existing telecommunication technology, and makes the exchange of bibliographic formats more convenient. MARC2 1 aims to redefine MARC format and improve its retrieval function to adapt to the network environment of 2 1 century. MARC 2 1 has five data formats: bibliographic format, authoritative format, collection format, classification format and community information format. At present, MARC 2 1 has been successfully applied to the British Library, the Library of Congress and the National Library of Canada. MARC2 1 allows UTF-8 to encode MARC-8 with variable-length characters using double bytes or Unicode. MARC-8 is a format based on ISO 2022, which can be used in Hebrew, Arabic, Greek and East Asian fonts.

MARC Standards Office is part of the Library of Congress.

MARC's recording structure is the implementation of ISO 2709, also known as ANSI/NISO Z39.2.