The American Physical Society publishes Physical Review A to E, Physical Review Letters, and Reviews of Modern Physics. All these journals will be available in electronic form from July 1st, 1997. Today APS offers both printed and electronic versions for Phys. Rev. A, C, D and Physical Review Letters.
The APS holds in high esteem its relationships with libraries. In fact, both the "Library Advisory Committee", along with the "Pricing Subcommittee", aim at promoting collaboration with libraries, as their institutions and researchers are the most important customers of the APS services.
As far as the costs of publication of the journals are concerned, there are three factors to be taken in account : editorial, composition, and printing and distribution (each about 1/3 of the total cost).

The number of articles received in 1995 was 20,939 and in 1996 there were 21,797. The ratio of published/received articles was 0.61 in 1995, 0.58 in 1996. Obviously, an increasing number of papers causes an increment in the costs and delays in publication. More than 90,000 pages were published, and a subscription to all the journals cost $.11 per page.
The articles submitted by American scientists accounted for 34 % of the total in 1994 and 29 % in 1996, while the papers submitted by European authors were 30.9 % in 1994 and 34.4 % in 1996. The APS has a database containing the names of 24,000 referees, of whom about 14,000 are used each year. Of these about one-third are from western European countries.
The Editorial Boards of the journals are made up of members from around the world.
APS is committed to making available electronic versions of all its print journals. Furthermore, it is working towards a paperless process of submission and refereeing, and therefore, requires suitable authoring tools of a high degree of sophistication. The increasing cost of "going online" is due to the need for both electronic and print technologies and personnel.
The APS is also beginning a process of decoupling electronic production from print production of the journals. In the future, the printed version should be a derivative of the electronic version, and not the opposite, as it is today.
If the Internet proves reliable enough, distribution of its journals will utilise this medium.
One might imagine a future where paper copies would be distributed electronically. A paper copy printed individually from the Web could cost up to 50 % less than the "traditional" paper version of the journal. At present, 2/3 of the cost of a journal is made up of composition, printing and distribution costs.
The APS is exploring the use of both mirror sites and dedicated electronic pipelines (jointly with other learned societies) for better distribution outside the U.S.
Its electronic publishing strategy includes:
- Plans for arranging worldwide distribution of the journals through the Internet.
- Enhancement of the capabilities of the electronic version. "Physical Review" articles will point to other sources of articles and abstracts, including the electronic preprint archives such as those hosted at LANL.
- Acceptation of "enhanced" papers (including multimedia features) for online publication.
- "Electronic only" publication for some articles.

For the future, APS intends to build up preprint archives together with other publishers. Moreover, it is establishing an archive of all "Phys. Rev." articles back to 1893, called the Physical Review On Line Archive (PROLA).
Articles back to 1985 are available in TIFF format and will eventually contain links to other Phys. Rev. articles in the references.
For the year 1998, the electronic version of the APS journals will be available without additional charge to institutional subscribers. The abstracts of the articles will be freely available to everybody, whereas access to articles and references will require a subscription. By July 1, 1997, Physical Review Letters and Physical Review C will contain links to other APS on line publications.
Finally, the APS recognizes two significant problems for all publishers of electronic journals: archival problems and the reliability of the Internet as a distribution medium for publications.