This collection is largely comprised of correspondence of the individuals and organizations who communicated with the editors of the Duke Mathematical Journal. All correspondence related to accepted papers was kept, as well as a significant amount of correspondence related to rejected papers before 1970. Referee reports are also included, as well as billing and correspondence between the Journal and various publishing houses. Topics covered include edits to papers, requests for offprint copies, correspondence between editors, correspondence with the American Mathematical Society, and the Journal's budget and funding. The collection ends with a series of record books recording each paper received and reviewed by the Journal.

The collection is organized into thirteen distinct series.

The first series is collated alphabetical files based upon the first letter of the subject, or the individual's last name. Each folder covers a year or a range of years, and each letter of the alphabet has five such folders, running from latest to earliest. Individual subject files for particularly prominent individuals and organizations are interspersed within the general alphabetical folders, and maintain roughly alphabetical order.

Series 2-7 are alphabetical files similar to the first series, with the difference that each letter of the alphabet recieved one folder covering a longer range of years, rather than several smaller folders for each letters. Subject files are interspersed alphabetically within series 2-3 as in series 1.

Series 8 is assorted subject files related to the period covered by series 4-7, where the subject files ceased to be interspersed throughout the series.

Series 9 represents a shift in the organizational structure, moving to a numeric system. As papers were submitted to the Journal, each was assigned a number and filed in chronological order, with any future correspondence related to that paper attached or stapled to the original piece of correspondence.

Series 10 retains the numerical structure of series 9, but adds two initial digits to each number signifying the year in which a paper was received.

Series 11 is also a numeric system, but the digits comprising the numbers changed in their significance. The first digit represents the last digit of the year (i.e. 1981 is 1, 1982 is 2, etc.). The next two digits signify the month the paper was received (01 represents January, 12 represents December). The following two digits represent the day of reception, while the last digit represents the order in which multiple papers were reviewed on that day. Thus, a paper with the number 104060 indicates that the paper was received on April 6th, 1981, and was the first paper to be processed that day.

Series 12 is comprised of miscellaneous correspondence unrelated to specific papers divided by year.

Series 13 is a series of record books documenting reception and processing of papers, and are in order by year.