Registering a systematic review is a workflow, not a form field you fill in once you already know what “PROSPERO” means — this guide walks through that workflow: when to register, what the record actually requires, how to submit it, and how to amend it once your review is underway. For a definition of PROSPERO itself and why prospective registration reduces duplication and outcome-switching bias, see the PROSPERO Registration dictionary entry; this page picks up where that one leaves off.
When to Register: The Timing Window
PROSPERO accepts registration at any point up to the point data extraction has been completed — but “accepted” is not the same as “useful.” The value of prospective registration comes from time-stamping your plan before you know the answer, so the practical target is registering before title/abstract screening begins, once your protocol is stable enough to commit to in writing. Registering after screening has started is still possible and still better than not registering at all, but reviewers and readers can reasonably discount how “prospective” a late registration really was.
Two consequences follow directly from this window:
- If your team is still negotiating the review question or eligibility criteria, finish that negotiation first — PROSPERO is not the place to draft a protocol from scratch.
- If you are already deep into data extraction, PROSPERO will not accept a new registration for that review. At that point the appropriate move is a transparent statement in the manuscript that the review was not prospectively registered, rather than attempting to force a late record through.
What to Have Ready Before You Open the Form
The registration form goes faster, and produces a more useful record, if the following are settled before you start typing rather than drafted live in the browser:
- Review question and framework. Most health-related reviews frame this with PICO or PICOTS (Population, Intervention, Comparator, Outcome, and — where relevant — Timing and Setting).
- Eligibility criteria. Study designs, populations, and outcomes that will be included or excluded, stated specifically enough that a second reviewer could apply them without asking you clarifying questions.
- Search strategy outline. Which databases and registers you will search (and, ideally, draft search strings) — see boolean search operators for literature searching if you haven’t finalized these yet.
- Planned approach to risk-of-bias/quality assessment and data synthesis. Which tool you’ll use to appraise included studies, and whether you plan a narrative synthesis, a meta-analysis, or both.
- Full review team, roles, and affiliations. Every named contributor, with ORCID IDs where available.
- Funding source and conflicts of interest. These are checked closely; leaving them vague is a common cause of a submission being sent back for revision.
Most of this overlaps deliberately with PRISMA-P, the reporting standard for systematic review protocols — if you’ve already drafted a PRISMA-P-structured protocol document, registering on PROSPERO is largely a matter of transferring it into the form rather than composing it from scratch. See PRISMA and systematic review methodology for how the protocol stage relates to the final reporting checklist.
The Registration Record: What the Form Covers
The PROSPERO form is organized into a small number of sections that map closely onto a PRISMA-P protocol. You do not need to memorize exact field labels — CRD periodically adjusts form wording — but every registration will need to address each of the categories below:
| Section | What it captures |
|---|---|
| Administrative details | Review title, anticipated start and completion dates, named review team members with affiliations and roles, corresponding author contact details, and any prior/related registrations. |
| Review question | The research question, stated using your framework (typically PICO/PICOTS), plus the condition or domain being studied. |
| Searching | Databases and registers to be searched, search date range, language and publication-status restrictions, and any planned supplementary search methods (citation chasing, grey literature, contacting authors). |
| Eligibility | Inclusion and exclusion criteria for participants, interventions/exposures, comparators, outcomes, and study designs. |
| Assessment methods | Data extraction approach, the risk-of-bias or quality-appraisal tool to be used, and how disagreements between reviewers will be resolved. |
| Synthesis plan | Whether results will be combined statistically (meta-analysis) or synthesized narratively, and any planned subgroup, sensitivity, or heterogeneity analyses — see heterogeneity in meta-analysis if this applies to your review. |
| Funding and conflicts of interest | Funding source(s) for the review itself, and any declared conflicts of interest among team members. These are mandatory and checked before publication. |
Step-by-Step: Submitting a New Registration
- Create a PROSPERO account. Registration and use of PROSPERO are free of charge. You will need an account to start, save, and submit a record.
- Start a new registration and select the correct record type. PROSPERO offers separate paths for a new systematic review versus an update to an existing review (relevant for living systematic reviews) — pick the one matching your situation before you start entering data, since switching later is not straightforward.
- Complete the administrative and review-question sections. Enter the title, team details, and your framed review question first — this is also where corresponding-author and funding details are captured.
- Enter your search strategy and eligibility criteria. Be as specific as your finalized protocol allows; vague eligibility criteria are one of the most common reasons a submission is returned for clarification before publication.
- Enter your assessment and synthesis plan. Name the specific risk-of-bias tool and describe, even briefly, how you intend to combine or synthesize results.
- Declare funding and conflicts of interest. These fields are mandatory; an incomplete declaration is a common hold-up.
- Save, review, and submit. PROSPERO lets you save a record as a draft before submitting, so it is worth having a second team member review the completed form against the protocol document before you submit it.
What Happens After You Submit
Following PROSPERO’s platform redesign, which launched 7 March 2025, records are automatically published once all named authors on the record have approved it — there is no separate manual CRD review-and-publish delay built into the standard path the way there sometimes was previously. Once published, your record receives a permanent, publicly citable identifier with a CRD42 prefix (for example, in the shape CRD42026123456). Save this number immediately — you will need it in your manuscript, in any funding or ethics reporting, and any time you come back to amend the record.
Amending a Registered Protocol
A systematic review protocol is a plan, not a contract signed in blood — deviations happen, and PROSPERO is built to accommodate that transparently rather than to penalize it. What matters is that a change is declared and dated, not that a review never deviates from its original protocol.
Typical reasons to amend a live record include:
- Adding a database or search source that wasn’t in the original search strategy
- Changing the risk-of-bias or quality-appraisal tool
- Revising an outcome definition or adding a secondary outcome
- Adding or removing review team members
- Updating the anticipated completion date
- Flagging that the review has now been submitted for, or published in, a journal
To amend a record, you sign in to your PROSPERO account, locate the registered review, and use the record’s amendment function to update the relevant fields — each amendment should include a brief explanation of what changed and why. PROSPERO retains a dated history of amendments rather than silently overwriting the original entry, which is precisely the point: a reader comparing your final manuscript to your registration should be able to see not just what the record says today, but what it said at each earlier stage and when it changed. A change that meaningfully alters what the review is answering — dropping a primary outcome, substantially narrowing eligibility criteria, or changing the comparator — is the kind of amendment worth calling out explicitly in the manuscript’s methods section, not just logging quietly on PROSPERO; reviewers checking your registration against your paper will look for exactly this kind of undisclosed drift.
How Journals Check Your Registration at Submission
PRISMA 2020 — the reporting checklist most systematic review and meta-analysis journals require or strongly recommend — includes a dedicated item (24a, in the “Other information” section) asking authors to provide registration information: the register name and registration number, or an explicit statement that the review was not registered. See PRISMA 2020 and PRISMA and systematic review methodology for the full checklist context.
In practice, this means two things at submission:
- Editors and peer reviewers commonly look up your CRD42 number and compare the registered protocol against the submitted manuscript — checking that the eligibility criteria, primary outcome, and planned analyses match, or that any differences are disclosed and explained rather than silently changed.
- Most journal submission systems and manuscript templates now ask for the registration number directly, typically in the abstract and again in the methods section, alongside a short statement of protocol availability.
This is also where a stale or unamended record becomes a liability: if your review’s actual eligibility criteria or outcomes shifted during conduct but the PROSPERO record was never updated to reflect it, a reviewer comparing the two will see an unexplained mismatch — which reads as a transparency problem even where the underlying change was reasonable.
Common Mistakes That Delay or Invalidate Registration
- Registering after data extraction is complete. PROSPERO’s window closes at that point; registering earlier — even mid-screening — is always better than not registering, but a review already through extraction cannot be prospectively registered at all.
- Vague eligibility criteria. “Studies of adults with diabetes” invites a request for clarification; “randomized controlled trials enrolling adults aged 18+ with a confirmed Type 2 diabetes diagnosis, excluding gestational diabetes” does not.
- Skipping the funding/conflicts-of-interest fields. These are mandatory, not optional, and an incomplete declaration is a routine reason a submission stalls.
- Registering a review type PROSPERO doesn’t cover. Scoping reviews and narrative/literature reviews are not eligible for PROSPERO, regardless of topic — see literature review vs. systematic review vs. scoping review if you’re not sure which category your project falls into, and systematic review vs. meta-analysis if the distinction between the two is the point of confusion.
- Letting the record go stale. Not updating the record after a genuine protocol deviation, then having the mismatch surface for the first time during peer review rather than being disclosed upfront.
Frequently Asked Questions
Is there a fee to register on PROSPERO?
No. Registration and use of PROSPERO are free of charge.
What review types are eligible for PROSPERO?
Systematic reviews with a health-related outcome — covering interventions, diagnosis, prognosis, aetiology, risk factors, genetic associations, or epidemiology — are eligible, and since 2021 this scope extends beyond clinical medicine into social care, welfare, public health, education, and criminal justice, provided the outcome is health-related. Scoping reviews and narrative/literature reviews are explicitly not eligible, regardless of subject matter.
Can I register after I’ve already started screening studies?
Yes, up until data extraction has been completed — but the earlier you register relative to screening, the more credibly “prospective” the record is. A review already past data extraction cannot register at all.
What does a CRD42 number look like, and where do I use it?
It’s a unique identifier assigned on publication of your record, in the shape CRD42 followed by the year and a sequence of digits (for example, CRD42026123456). Cite it in your manuscript’s abstract and methods section, and in any funder or ethics-committee reporting that asks for a registration reference.
Do I need to register on PROSPERO if my journal doesn’t explicitly require it?
Registration is not universally mandated the way clinical trial registration is under ICMJE and FDAAA rules, but PRISMA 2020 (checklist item 24a) asks every compliant review to report a registration number or explicitly state that the review was not registered — so an unregistered review still has to say so, in writing, in the published paper. Registering avoids that disclosure and the credibility questions it can invite.
What’s the difference between amending a record and it simply going out of date?
An amendment is a deliberate, dated, explained update you make through PROSPERO’s own record-editing function, and it stays visible in the record’s history. A record going “out of date” means the review’s actual conduct changed but nobody updated PROSPERO to reflect it — that’s the situation that creates an unexplained mismatch for peer reviewers to catch later.
For related process context, see the EQUATOR Network reporting guidelines for choosing the right reporting checklist beyond PRISMA, the Cochrane Handbook for Systematic Reviews of Interventions for conduct-level methodology, and pre-registration for how the same prospective-transparency logic applies to primary studies rather than syntheses.
