Home › Guides › ROR research organization data

ROR ID lookup: match messy institution names to one canonical record

The Research Organization Registry (ROR) is the community-run, open registry of research organizations — universities, institutes, hospitals, companies and funders. This actor searches it by name, ROR ID or website domain, optionally filtered by organization type, and returns one clean row per organization: canonical name, ROR ID, country and city, coordinates, website, Wikipedia page, founding year, and its place in a parent and child hierarchy.

Who it is for

• Data teams deduplicating institution names across datasets.
• Research-intelligence analysts who need a canonical ID to join grant, publication or trial records on.
• Research-sales teams looking for the parent university behind a research institute.
• Anyone building a clean list of organizations of one type in one country.

What you get

One row per organization. Fill rates are from a measured run of 31 rows for 4 queries on 2026-09-20:

A real row from the default run, shortened:

{
  "companyName": "Siemens (Germany)",
  "registrationNumber": "https://ror.org/059mq0909",
  "status": "active",
  "legalForm": "company, funder",
  "establishedYear": 1847,
  "city": "Munich",
  "countryName": "Germany",
  "countryCode": "DE",
  "latitude": 48.13743,
  "longitude": 11.57549,
  "website": "http://www.siemens.com/entry/cc/en/",
  "wikipediaUrl": "https://en.wikipedia.org/wiki/Siemens"
}

Use cases

• Entity resolution. Turn a column of messy institution names into canonical ROR IDs.
• Data joining. ROR IDs are the standard key across grant, publication and trial datasets.
• Hierarchy mapping. Find the parent university behind an institute, or all children of a system.
• Geographic analysis. Coordinates on every row, so organizations plot straight onto a map.
• Cross-registry linking. externalIds carries GRID, ISNI, Wikidata and Fundref identifiers.

How to run it

Put anything you have into queries — a name, a ROR ID or a domain. The prefilled input shows all three at once:

{
  "queries": ["Siemens", "Max Planck Society", "059mq0909", "novartis.com"],
  "maxResultsPerQuery": 10
}

• queries — organization names, ROR IDs (bare or as a full URL) or website domains, mixed freely.
• organizationType — optional filter: company, education, healthcare, government, facility, nonprofit, funder, archive or other.
• maxResultsPerQuery — caps how many organizations come back per query.

The measured run above returned 31 rows in 5 seconds. The actor is also on the Apify MCP server as foxlabs/ror-research-organization-data, and it supports agentic payments: an AI agent can run it and settle the bill over x402 (USDC on Base) or Skyfire, without an Apify account of its own.

Honest limits

• Hierarchy, alias and domain fields are sparse by nature. ROR records vary in depth; the fill rates above say exactly how sparse. Use website, which is filled on every row, rather than domains.
• ROR’s hierarchy is not corporate ownership. Parent and child links are ROR’s own structure: a university hospital can be a ROR child of the university without being a subsidiary.
• Research organizations only. ROR registers organizations involved in research, so a company with no research activity may simply not be there — absence says nothing about the company.
• No registry or contact fields. ROR does not publish tax numbers, street addresses, industry codes, headcount, capital or contact details, so those columns were removed rather than shipped empty.
• Branches by country. Large institutions have national branches in ROR, such as Siemens (Germany). Pick by countryCode, or search the ROR ID directly.
• Precise queries return one row. A ROR ID or domain resolves to one organization by design. ROR IDs are checksummed, so check for typos if one returns nothing.

Where the data comes from

ROR publishes its IDs and metadata under a CC0 public-domain dedication, explicitly for reuse, and states there are no restrictions on access to or use of ROR IDs and metadata. The rows are institutional records, not personal data. The actor charges for running the extraction and delivering clean rows, not for the data itself. It is an independent tool, not affiliated with or endorsed by ROR.

Related data and guides

• Browse the research funding and institution datasets on the home page.
• Pharma pipeline data — why resolving institutions to a stable ID matters when you join grants, trials and approvals.
• SEC filings, enforcement and the LEI — why identifiers beat name matching, on the company side.
• Company ID crosswalk — LEI, ISIN, CIK and VAT for companies — and the NIH and NSF funding guides.

Frequently asked questions

Can I look up a ROR ID by website domain?

Yes. Put the domain, such as novartis.com, in queries. Names and ROR IDs work in the same field, so one list can mix all three.

Is ROR data free to reuse?

ROR IDs and metadata are released under CC0. The actor charges for running the extraction and delivering clean rows, not for the data itself.

Does ROR include companies?

Yes. ROR covers companies and funders alongside universities and hospitals; the default run resolves Siemens and Novartis.

Why is the domains field empty on most rows?

ROR fills domains for a minority of records — 13% of rows in the measured run. The website field is filled on every row and is the reliable one.

How do I list every organization of one type in a country?

Search a broad term with organizationType set, then filter the output by countryCode.