What is OJS (Open Journal Systems) and how does it work?

What is OJS used for, and by whom

According to the usage data PKP publishes, 58,221 journals used OJS to publish five or more items in 2025, in some 156 countries, and more than 11 million items have been published with OJS since it was created. PKP describes it as the most widely used end-to-end scholarly publishing platform, and it is especially common among diamond open-access journals, which charge neither authors nor readers.

Three characteristics explain that reach. OJS is free software under the GNU General Public License, so there is no license fee per journal, per article or per user. A single installation can run one journal or many, each with its own editorial team, policies, sections, email templates and look, which is why university presses often publish their whole portfolio on one OJS. And the project is actively maintained: the current long-term support (LTS) release is OJS 3.5, which PKP supports until the second quarter of 2029. Our OJS setup and hosting guide covers versions and requirements in detail.

The problem it solves: running a journal by email

Before adopting a system like OJS, most scholarly journals, especially university journals, run their editorial workflow with a combination of email, shared folders and a spreadsheet where someone records the status of each manuscript. That works while the journal receives few articles a year. It stops working as soon as the journal grows.

The symptoms are always the same. The editor does not know for sure how many manuscripts are awaiting review or how long they have been waiting. Reviewers receive files by email and respond, or do not, without leaving a record the rest of the team can see. Manuscript versions pile up as attachments with different names, and nobody can be sure which one is final. And when an audit, a funding call or an indexing application arrives, reconstructing how long each article took from submission to publication means reading emails one by one.

OJS does not make peer review faster, because that depends on people. What it does is make the whole process visible and auditable: who has the manuscript, at what stage, what is missing and since when. That visibility is the real difference between managing a journal and chasing after it.

The four stages of the OJS editorial workflow

OJS organizes each manuscript into four sequential stages, each with its own panel, its own files and its own internal discussion tool:

1. Submission. The author uploads the manuscript through a step-by-step wizard that reminds them of the journal's guidelines and captures the basic metadata (title, abstract, keywords, authorship) from the start. Here the editor decides whether the manuscript is declined outright, assigned to a section editor or sent to review.
2. Review. The peer review stage, including any revisions the author must make in response, before the decision to publish is taken.
3. Copyediting. The accepted manuscript goes through language, style and format editing, typically with an initial copyedit followed by the author's review of those changes.
4. Production. The edited files are turned into the final publication formats, such as PDF and HTML, known as galleys. They get a final check and the article is ready to be scheduled into an issue.

Each stage has its own discussion thread, where editor, reviewers, author and copyeditors talk about that specific manuscript, with email notifications to those involved. The conversation is archived as part of the article's editorial history, something no chain of loose emails can offer.

There is also an alternative route, the Quick Submit plugin, for publishing content that does not need to go through the full workflow, for example when loading back issues already published in another system. It is not meant for the regular editorial flow of new articles.

Email and spreadsheets vs. OJS

Roles and permissions in OJS

OJS assigns permissions by role and by workflow stage, so each person sees and does only what corresponds to them:

• Journal manager. Configures the journal: sections, policies, review forms, email templates and users. It is the administrative role, not necessarily the one deciding on content.
• Editor and section editor. Receive submissions, assign reviewers, make editorial decisions and move manuscripts between stages.
• Reviewer. Accesses only the manuscript assigned to them, sees the journal's review form and submits a recommendation inside the system.
• Author. Submits the manuscript, responds to revision requests and approves the final galleys before publication.
• Copyeditor and layout editor. Work in the copyediting and production stages, with access limited to those files.

Each role can have independent permissions on the four stages, which allows, for example, an external reviewer to enter only the review stage without seeing anything else in the editorial process.

What OJS connects to: DOIs, ORCID, indexing and preservation

A journal does not live in isolation, and much of the value of OJS lies in the standards it plugs into:

• DOIs through Crossref, with registration and automatic deposit from OJS; DataCite is supported as well. The journal still needs Crossref membership, directly or through a sponsor, and its own DOI prefix.
• ORCID authentication for authors and ROR identifiers for affiliations.
• OAI-PMH and Dublin Core metadata, so aggregators, discovery services and Google Scholar can harvest and index the journal, plus export to DOAJ.
• Preservation through archiving services such as the PKP Preservation Network (PKP PN) and CLOCKSS/LOCKSS, a requirement several indexes value.
• Multilingual metadata: one article can carry its title, abstract and keywords in several languages, as many regional indexes require.

These connections matter for funder policies too. Plan S, for example, expects compliant open-access journals to be registered in DOAJ, use persistent identifiers, deposit content with a long-term preservation program and publish high-quality metadata under a CC0 dedication, without embargoes. OJS can carry the technical side of several of these requirements; the editorial and licensing side belongs to the journal. Full-text XML production in JATS, if the journal needs it, is a separate editorial service rather than part of the software implementation.

What OJS is not

It is worth being precise about the limits of the system, because overselling creates expectations that later disappoint. OJS does not evaluate the scientific quality of a manuscript or replace editorial judgment: it structures the process, but the decision to accept or reject remains human. It does not, on its own, get a journal indexed in DOAJ, Scopus, Web of Science or regional indexes such as SciELO and Latindex, nor register DOIs without a Crossref membership. Those are integrations and procedures configured on top of OJS, and they depend largely on the journal meeting editorial criteria the software cannot guarantee, such as publication frequency, an editorial board and an open-access policy.

And OJS is not an institutional repository. If the university also needs a home for theses, reports and other non-serial output, that is a different platform, typically DSpace, although the two systems coexist well and share identifiers such as DOIs.

In BiblioLabTech's projects, the first conversation with a journal almost always starts by correcting this expectation: OJS organizes the editorial process; it does not fully automate it or replace the editorial board. What it does eliminate, measurably, is the time lost administering the process instead of editing.

Which journals should use OJS

OJS makes sense for any scholarly, technical or cultural journal that publishes with some regularity and has, or wants to have, formal peer review: faculty journals, research group journals, scholarly society journals and university presses. It also suits journals that currently publish on a proprietary system or a custom-built site and want to stop depending on an external developer for every change, because OJS is free software and the installation stays under the institution's control.

A publication that appears sporadically without a defined peer review process does not need it, at least not yet. There the problem is not the system but the absence of an editorial process, and automating an undefined process solves nothing. When BiblioLabTech implements OJS for a university, the first thing we review with the editorial team is exactly that: whether policies and workflow are already defined on paper, or need to be defined before they are configured in the system.

Next step

Understanding what OJS is comes first. The second step is understanding how it is put into service in practice, what you need, in what order, where to host it and how long it realistically takes, which is what our OJS setup and hosting guide covers. If your institution also runs a DSpace repository, read about DSpace-CRIS as well, because articles published in OJS with DOIs are often the same outputs that must later be reported by researcher and by project. BiblioLabTech implements OJS through its OJS implementation and support service, with the same team that runs DSpace repositories and Koha catalogs for universities.

OJS is free software developed by the Public Knowledge Project.