ClinicalTrials.gov Search — Trials, Sponsors, Sites and Eligibility

Search the US registry of clinical studies — 603 494 records — by condition, drug, sponsor, phase, status and country, and get back flat rows instead of nested JSON.

The registry's own API returns each study as a tree of protocol modules nested five levels deep. This Actor searches it and flattens the result: one row per trial, with the sponsor, the phase, the enrollment, the eligibility, the primary outcome and the sites in named columns you can sort in a spreadsheet.

No API key. Public NIH data.

What a search returns

Input: condition glioblastoma, status RECRUITING, phases PHASE2 and PHASE3, type INTERVENTIONAL, country United States.

The summary row reports 73 matching, 60 returned (the cap set in the input), and the filters that were applied. Then one trial row each. One of them, NCT06556563:

Field	Value
briefTitle	EF-41/KEYNOTE D58: Phase 3 Study of Optune Concomitant With Temozolomide Plus Pembrolizumab in Newly Diagnosed Glioblastoma
leadSponsor, sponsorClass	NovoCure GmbH · INDUSTRY
collaborators	Merck Sharp & Dohme LLC
phase, studyType	Phase 3 · INTERVENTIONAL
allocation, masking	RANDOMIZED · QUADRUPLE
enrollmentCount, enrollmentType	741 · ESTIMATED
startDate, primaryCompletionDate	2025-02-03 (ACTUAL) · 2029-04 (ESTIMATED)
primaryOutcome, primaryOutcomeTimeFrame	Overall survival · 24 months
interventionNames	Optune® device; Temozolomide; Pembrolizumab; Placebo
locationCount, countryCount	102 · 12
recruitingSiteCount	19
contactEmail, contactPhone	clinicaltrials@novocure.com · 1-877-678-8611
minimumAge, sex, healthyVolunteers	18 Years · ALL · false
url	https://clinicaltrials.gov/study/NCT06556563

Plus eligibilityCriteria — the sponsor's full inclusion and exclusion list — briefSummary, meshTerms, the MeSH vocabulary the registry assigned, and a locations array with facility, city, state, country, per-site recruitment status and latitude/longitude.

Two numbers worth reading twice

locationCount is the number of sites the trial lists. recruitingSiteCount is how many of them are actually open. On the trial above they are 102 and 19: the trial's overall status is RECRUITING, but most of its individual sites have already closed enrollment. Filtering on the trial status alone will send a patient to a hospital that stopped taking people last year.

primaryCompletionDate is when the primary endpoint is read out. completionDate is when the last follow-up ends, often years later. For read-out timing the first is the one you want.

Sponsor search matches whole words

Moderna returns 1 trial. ModernaTX returns 142. The registry spells the company ModernaTX, Inc., and the search matches whole words, not prefixes — so the single hit for Moderna is a Vertex trial whose collaborator is literally named "Moderna, Inc".

Extra words narrow rather than widen: Pfizer returns 6 085 trials, Pfizer Inc returns 705.

The search also covers collaborators, not only the lead sponsor. Of the 142 ModernaTX trials, 114 are led by ModernaTX and 28 are academic or partner trials it co-funds — a Merck trial, three NIAID trials, three from the International AIDS Vaccine Initiative. Both leadSponsor and collaborators are on every row, so the distinction is visible rather than buried. Run a small search first and look at the leadSponsor column before scaling up.

Status is where the interesting records are

The 142 ModernaTX trials break down as 85 COMPLETED, 21 ACTIVE_NOT_RECRUITING, 18 RECRUITING, 9 TERMINATED, 5 WITHDRAWN, 2 NOT_YET_RECRUITING and 2 UNKNOWN.

TERMINATED and WITHDRAWN are the programmes that stopped, with the dates attached. UNKNOWN means the sponsor has not updated the record past its own expected completion date, which is its own signal. The Currently recruiting only switch is shorthand for RECRUITING plus NOT_YET_RECRUITING, and it overrides the status list.

Phases

A trial can declare two phases at once. Searching glioblastoma for PHASE2 and PHASE3 returned 30 trials marked Phase 2, 22 marked Phase 1/Phase 2, 7 Phase 3 and 1 Phase 2/Phase 3 — asking for a phase returns the combined designations that include it, which is what you want. NA is not a missing value; it is the phase of trials that have none, which covers most device, behavioural and observational work.

Who uses this

Clinical operations and feasibility teams sizing a site list before a study starts, and checking who else is recruiting the same patients in the same country. Competitive intelligence and business development at pharma and biotech, watching a sponsor's pipeline by first-posted date. Patient advocacy groups and trial-matching services building searchable lists of open studies with contact details. Investors and analysts tracking read-out timing from primaryCompletionDate. Academic groups doing systematic reviews and registry-based research, where the terminated and withdrawn records matter as much as the completed ones.

Scope and limits

Registration, not results. This Actor returns the protocol record: what the trial says it will do, who runs it and where. It does not return posted results, adverse-event tables or publications.

Sponsors write their own records. Condition names, keywords and intervention names are the sponsor's wording, not a controlled vocabulary. That is why the registry's synonym expansion matters, and why a search for type 2 diabetes also returns trials that wrote T2DM. The meshTerms column carries the standardised vocabulary the registry assigns on top.

Records go stale. lastUpdateDate is on every row. A trial marked RECRUITING whose last update is three years old is not reliable evidence that it is still recruiting.

Not every trial is here. ClinicalTrials.gov is comprehensive for US-regulated studies and voluntary elsewhere. A trial run only in the EU may appear in the EU register instead, and a trial run only in China or Japan may appear only in its national registry.

Source and licence

ClinicalTrials.gov, the registry maintained by the US National Library of Medicine at the National Institutes of Health, through its public v2 API at clinicaltrials.gov/api/v2/studies. A US federal government work, public domain, no key and no registration. Queried live on every run, so the results are the registry as it stands at the moment of the run, not a cached copy.

Frequently asked

Why did my sponsor search return almost nothing? Because it matches whole words. Try the company's registered name as the registry spells it — ModernaTX rather than Moderna. Search the sponsor on its own with no other filters first, then read the leadSponsor column to see which spelling the registry actually uses before you add filters.

Does a country filter mean the trial only runs there? No. It means at least one site is in that country. The row carries countries and countryCount so you can see the rest — the glioblastoma trial above matched on the United States and also runs in eleven other countries.

Why is phase empty on some rows? Observational studies have no phase at all and the field is absent. Interventional studies without one carry the explicit value NA.

How many trials can one run return? Up to ten thousand. Results are paged 200 at a time against the registry's cursor-based pagination, which was checked over four consecutive pages: 700 trials, 700 distinct NCT IDs, no overlap.

Is the eligibility text complete? Yes, up to 20 000 characters, which covers essentially every record. Longer ones are marked with a trailing ellipsis, and the full text is always at the url on the row.

Why does a search with only filters and no search term refuse to run? Because a filter alone would return an arbitrary slice of 603 494 studies sorted by relevance to nothing. Give at least one condition, intervention, sponsor or free-text term.