Your university has a repository. So does your field. Here's what each does well — and why the answer is often both.
Two Different Kinds of Archive
When researchers talk about depositing work in a repository, they usually mean one of two quite different things. An institutional repository is run by a university or research organisation and exists to gather everything produced under that roof — papers, theses, datasets, technical reports, even grey literature. A subject repository is organised around a discipline rather than an address, open to anyone working in that field regardless of where they're employed. Both are legitimate, permanent homes for scholarship. They just have different logics.
Institutional repositories exist because universities have a stake in demonstrating their own research output. They serve compliance functions — funding bodies and national assessment frameworks often require institutions to track and expose their work — and they preserve the full breadth of what a university produces, not just journal articles. A doctoral thesis, a dataset underpinning a clinical trial, a set of teaching resources: all of these belong somewhere, and an institutional repository is built to hold the lot. The metadata is structured to serve internal reporting as much as external discovery, and long-term preservation is usually a formal commitment.
Subject repositories have a narrower but arguably sharper focus. arXiv, which has hosted preprints in physics, mathematics and adjacent fields since 1991, is the canonical example: researchers submit there because that is where their community reads and discusses work, often before it is formally published. PubMed Central serves biomedicine. Zenodo, hosted at CERN, operates as a cross-disciplinary catch-all but is heavily used by research communities that lack a dedicated subject home. The Directory of Open Access Journals (DOAJ) doesn't host preprints itself, but it indexes the open journals many subject repositories feed into. What these platforms share is community gravity: deposit there because that is where your peers are already looking.
Strengths, Gaps and the Case for Both
The practical difference comes down to discoverability versus comprehensiveness. Subject repositories are where working researchers in a field habitually search; depositing in arXiv means a particle physicist in Geneva sees your work the morning it goes up, not six months later when it clears peer review and appears behind a paywall. Institutional repositories, by contrast, aggregate everything from a single institution into one searchable record — useful for administrators, accreditation processes, and anyone trying to understand what a particular university actually produces.
Neither type is without limitations. Institutional repositories vary enormously in quality: a well-staffed research library will maintain rich metadata, regular harvesting by Google Scholar, and robust preservation infrastructure; a neglected one may hold files that are hard to find and poorly described. Subject repositories, meanwhile, tend to serve established disciplines better than newer or interdisciplinary fields — if your research sits at the intersection of law and machine learning, it may not have an obvious subject home at all. Zenodo was built partly to fill that gap, offering persistent identifiers and Creative Commons licensing for work that doesn't fit neatly elsewhere.
The good news is that depositing in one repository rarely precludes depositing in the other. Many universities actively encourage — or mandate — deposit into the institutional repository while recognising that researchers will also post to arXiv or a disciplinary server. The two deposits typically serve different audiences and different purposes without conflicting. The version question matters here: what you put in each repository should be appropriate — a preprint before peer review, an accepted manuscript after — but the principle of dual deposit is widely accepted. If you want a fuller picture of version rules and licences, the practical steps are covered in the guide to depositing your own work.
One underappreciated advantage of subject repositories is speed. In fast-moving fields, the months between submission and formal publication are not trivial; a preprint on arXiv is citable, readable and discoverable within days. Institutional repositories, which often wait for a published or accepted version before deposit, operate on a slower clock. For researchers, the choice is less either/or than it first appears: post early to the subject repository so your community can engage with the work, and fulfil your institutional obligation by depositing the accepted version when the journal confirms what is permitted.