Slovak Academy of Sciences (SAS) is publishing approximately 50 scientific journals. Others are publishid by universities. We will concentrate ourselves on mathematics. Four of the above mentioned journals are devoted to mathematics and computer science. The SAS supports conversion of the articles into PostScript (PS) format and publishing of the journals in paper and electronic form simultaneously. Papers in PS format are stored on CD ROMs and on file server savba.sk. On the home page http://savba.sk there are several preprints stored too. We propose to publish all our traditional journals in both paper and electronic form. Preparation of documents is cheaper only in the step of the electronic type-setting. Most authors are able to prepare their documents in TeX, which is the broadly accepted language for describing or formatting the documents in mathematics. There are different requirements for the pre print or not-refereed documents and electronic journal. Electronic journal must fulfill scientific standard, i.e. Editorial Board has to assure that accepted articles are of good scientific quality. This is possible to achieve by refereeing each article by experts. In mathematics for example we have only two journals on CD ROMs, both are abstracted in Zentralblatt fur Mathematik and Mathematical Reviews. We have got these journals due to our cooperation with publishers. Slovak academic institutions (i.e. universities and institutes of the SAS) are connected to the ERBONE (European Backbone) system in Bratislava by 1 Mb line. In mathematics, we have approached the European Mathematical Information Service (EMIS) in Berlin, through the Internet. EMIS is mirrored on 16 further servers in Europe and 10 servers outside Europe. There are 22 journals stored in the electronic library of EMIS. Some of them are available in this library only one or two years after publication. We will store all our mathematical journals in this EMIS electronic library in near future. Our experience shows that releasing the older volumes of some scientific journals for free approach in some database does not decrease the number of subscribers. We believe that it would be useful to propose to great publishers such as Springer-Verlag, Academic Press, Elsevier, M.Dekker, . . . to consider releasing older volumes of scientific journals and proceedings for storing and free access on specialized servers. Except of databases with refereed journals it will be useful to establish regional databases with preprints and not-refereed articles in various branches of science as it is used in physics.