PROSPERO registration is free for systematic reviews with outcomes directly relevant to human health. Search the register for similar reviews, finalise your protocol, then complete the online form before data extraction starts. Once the senior guarantor and every team member have confirmed, the record is published with a CRD number, which you report under PRISMA 2020 item 24a.
PROSPERO is an international registry of systematic review protocols, run by the Centre for Reviews and Dissemination (CRD) at the University of York. A record is a short public summary of your plan, not a full protocol or a formal publication. Whether to register at all is covered in our comparison of systematic reviews, meta-analyses and literature reviews.
Confirm your review is eligible before PROSPERO registration
PROSPERO accepts systematic reviews with outcomes of direct relevance to human health. Reviews of animal studies relevant to human health go to PROSPERO4animals, on the same platform. At the time of writing, only the intervention-review form is live, and PROSPERO asks every review type to use it. The table combines PROSPERO's published eligibility rules with its registration guidance and FAQs.
| Your review | Accepted? | What to do |
|---|---|---|
| Human-health outcomes, including a living review | Yes | Register; select the living review option if it applies |
| Animal studies relevant to human health | Yes, on PROSPERO4animals | Human and animal studies together need two records, each noting the link |
| Data extraction started, or some individual participant data (IPD) already held | No | Report the review as unregistered |
| Indirect health outcomes, sporting performance, evidence and gap maps, reviews of clinical practice guidelines, in-vitro studies only, or a literature review whose only systematic step is the search | No | Lodge the protocol elsewhere, such as the Open Science Framework (OSF) |
| Scoping review | Not currently | Register the protocol on OSF |
| Student module or training review | Never published | Choose the training option; the record can be shared with supervisors but never appears in searches |
For reviews outside its scope, PROSPERO itself points to OSF Registries, where registration is free. Our guide to preregistration explains what a timestamped public plan protects against.
Search PROSPERO for similar reviews before you fix the question
PROSPERO's public listing lets researchers check for an existing review before planning a new one. Search the register with your condition, intervention and outcome terms first.
While you fill in the form, PROSPERO looks for registered reviews that resemble yours, matching on the title, the review question, your keywords and the labels you chose for the health condition, the intervention and any comparator. If it finds similar reviews, it does not stop you. It asks you to explain why the new review is needed, though an answer is optional, and your reasons become part of the public record. Make the same case in the Rationale field, which asks you to look beyond PROSPERO as well.
Register after the protocol is final and before data extraction starts
PROSPERO's cut-off is the start of data extraction or, for an individual participant data (IPD) review, the receipt of any participant data. On the form's "Current stage of the review" table, pilot work, searching and screening may be under way. Data extraction, risk of bias assessment and data synthesis must all be marked Not started, or a first registration is refused.
That cut-off is the latest point to register, not the recommended one. PROSPERO expects the protocol to be finalised, or very nearly, with the time or funding to finish in place. Give true dates and stages: PROSPERO treats knowingly false ones as research misconduct and may withdraw such records.
If you are already past the cut-off, do not register; report the review as unregistered. PROSPERO tells authors never to register a completed review, even when a journal asks for a number. It also tells journals that registration after submission is too late. Our guide to clinical trial registration makes the case for open disclosure; its advice to register late applies to trials, not to PROSPERO.
Fill every field, even if you upload a full protocol
The record owner needs a PROSPERO account and an ORCID iD. The team needs at least two members. One must be the guarantor, the senior team member who approves the record's content. A PDF of the full protocol may be uploaded, but PROSPERO will not accept a record that points readers to it instead of completing the fields. The form also flags a field answered only "N/A", "None" or "Not applicable" as not valid. Full protocols are reported with PRISMA-P 2015. At the time of writing (October 2026), the form's sections ask for the following.
| Form section | What to have ready |
|---|---|
| Review title and basic details | Working title in English; condition or domain; rationale; objectives; keywords; country |
| Eligibility criteria | Population, intervention or exposure, comparator, study designs and context, with inclusion and exclusion criteria |
| Timeline of the review | Start and end dates; the living review option |
| Availability of full protocol | An optional PDF |
| Searching and screening | Main databases; unpublished sources; language and date limits; an optional search strategy PDF; the selection process |
| Data collection process | Data to extract; number of reviewers and how disagreements are resolved; risk of bias tool and level; reporting bias; certainty assessment |
| Outcomes to be analysed | Main and additional outcomes, each with definition, instrument, time points and effect measure |
| Planned data synthesis | Meta-analysis model, heterogeneity, missing data, subgroup analyses with rationale, sensitivity analyses |
| Current review stage | The stage table, answered truthfully |
| Review affiliation, funding and peer review | Team members, affiliation and funding source |
| Additional information | Conflicts of interest, Medical Subject Headings, notes linking related records |
What happens after you submit the record
PROSPERO's registration help page sets out the sequence. Automated checks run as you enter each field. The guarantor approves the content, then every named team member confirms their participation and consent to publish their name and email. A number is then assigned and the record is published immediately.
Publication is not a quality check. PROSPERO does not check content beyond those automated checks, does not peer review it and does not endorse it, so avoid wording in your paper that implies otherwise. Random checks may still lead to the withdrawal of out-of-scope or inaccurate records. A CRD number is the registration number PROSPERO assigns to each published record: "CRD" followed by digits, with no fixed length. This guide writes it as CRD[number].
Amend the record and update its status as the review moves on
A Minor amendment to a published PROSPERO record covers only a few fields. Most changes, especially to methods or the review team, need a Major revision. Each edit creates a new version. A Major revision is published only after every team member approves it, while a Minor amendment needs no team approval. A dated audit trail joins the record, and earlier versions stay public.
When the planned end date arrives, the record owner is emailed to update it. Mark the review completed, or discontinued with the reason, and add the citation once the paper is published. In the paper, report each deviation with its date and reason, as our guide to writing a research protocol describes.
Report the CRD number under PRISMA 2020 item 24
PRISMA 2020 asks you to report registration status, not to register. Item 24 of its 27-item checklist covers "Registration and protocol", and item 12 of the 12-item checklist for abstracts covers registration. Check each line against the PRISMA 2020 checklist.
| Item | What PRISMA 2020 asks for |
|---|---|
| Abstracts item 12 | Register name and registration number |
| 24a Registration | Register name and registration number, or a statement that the review was not registered |
| 24b Protocol | Where the protocol can be accessed, such as a citation, DOI or link, or that none was prepared |
| 24c Amendments | Each amendment to the registered information or protocol, its reason, and the review stage when it was made |
Copy the statements below and keep only the version that is true.
Abstract: Registration: PROSPERO CRD[number].
Methods, registered review: This review was registered in PROSPERO (CRD[number]) on [date], before data extraction began. The full protocol is available at [citation or link]. / No protocol was prepared beyond the PROSPERO record.
Methods, amendments: After registration, we [change] at the [review stage] stage because [reason]. The change is recorded in the PROSPERO version history.
Methods, unregistered review: This review was not registered. The protocol is available at [citation or link]. / No protocol was prepared.
Reference: [author], [author]. [title]. PROSPERO [year] CRD[number]. Available from https://www.crd.york.ac.uk/PROSPERO/view/CRD[number].
Run this pre-registration checklist before you submit
- Outcomes are directly relevant to human health, and the review is not an excluded type.
- PROSPERO and the literature have been searched for similar reviews, and the need is explained.
- The protocol is final or nearly so, with time or funding to finish.
- Data extraction, risk of bias assessment and synthesis have not started; no IPD have arrived.
- The owner has a PROSPERO account and ORCID iD, and at least two members, including a senior guarantor, have agreed to be named.
- Every field is complete and in English, even if you upload a protocol, which may be in any language.
- The start date and review stage are true on the day you submit.
At Directive Publications, we encourage, but do not require, protocol registration for systematic reviews, for example in PROSPERO, with the registration number stated in the abstract. Our page on Types of Articles We Publish sets this out. The same page requires a completed PRISMA checklist and flow diagram for systematic reviews and meta-analyses. If you spot an error in this guide, here is how to report a problem to us.