Do I have to register my systematic review protocol?

Not legally, but functionally yes for any review you intend to publish in a peer-reviewed journal. Most major journals (BMJ, JAMA, Cochrane Library, JCE, BMJ EBM) either require or strongly recommend prospective registration. PRISMA 2020 also expects a registration entry to appear in the Methods section. Unregistered reviews are accepted, but they face higher scrutiny on selective reporting.

What's the practical difference between PROSPERO and OSF?

PROSPERO is a curated, gated registry — submissions are screened by NIHR staff and rejected if they don't meet eligibility (most commonly: scoping reviews, in-progress data extraction, methodology reviews). OSF is a platform — anyone can register a project in minutes with no eligibility check. PROSPERO is the higher-trust signal for clinical reviews; OSF is the broader-scope, faster, eligibility-flexible alternative.

How long does PROSPERO registration actually take in 2026?

PROSPERO publishes a target turnaround of around 6 weeks; the actual queue varies. Most submissions clear in 4–10 weeks, with peaks of 12+ weeks during high-volume periods. Plan registration as the second step of the protocol phase, not the last — submit when the protocol is locked, then wait for confirmation while the team builds the search strategy.

Can I register a scoping review on PROSPERO?

Generally no. PROSPERO's stated scope is reviews where there is a health-related outcome and the review's purpose is to inform clinical practice or policy. Scoping reviews, mapping reviews, and methodology reviews are typically out of scope. The OSF Registries platform — and increasingly the JBI scoping review template — are the standard alternatives. See our guide to scoping reviews for the methodology context.

What if I've already started screening — can I still register?

PROSPERO requires registration before data extraction begins (not before screening). If you have completed screening but not extraction, you can still register on PROSPERO, though the form will ask for your screening start date. If you have started extraction, registration is no longer prospective and most journals will treat the review as retrospectively registered, which is a weaker signal but still better than no registration.

Where to Register a Systematic Review Protocol: PROSPERO vs OSF vs Cochrane Protocols

Registering a systematic review protocol used to be a niche practice. By 2026, it is a baseline expectation. PRISMA 2020 lists registration as a required reporting item; the major medical journals require or strongly recommend it; and the case law on selective-outcome reporting has grown substantial enough that an unregistered review now invites the question "what changed between protocol and publication?"

The harder question is where to register. Three registries dominate, and the choice is not interchangeable.

The three registries, summarized

Dimension	PROSPERO	OSF Registries	Cochrane Protocols
Operator	NIHR Centre for Reviews and Dissemination (UK)	Center for Open Science (US)	Cochrane Collaboration
Scope	Reviews with health-related outcomes, intent to inform practice/policy	Any research project; no eligibility gate	Reviews accepted into a Cochrane Group's editorial process
Curation	Editorial check; rejections common	None — self-service	Full editorial peer review of the protocol
Cost	Free	Free	Free; subject to Cochrane Group acceptance
Turnaround	4–10+ weeks	Minutes	Months (peer review)
Output format	Public registry entry, citable ID	Public registered project, DOI	Published protocol in Cochrane Library
Amendment policy	Versioned; full history visible	Versioned; full history visible	Versioned with editorial review
PRISMA 2020 recognition	Yes	Yes	Yes

The differences look small in a table. They are not small in practice.

When to choose PROSPERO

PROSPERO (International Prospective Register of Systematic Reviews) is operated by the UK's National Institute for Health and Care Research, Centre for Reviews and Dissemination, since 2011. It is the registry most clinical journals expect to see in your methods section.

It is the right choice when all of the following are true.

The review's primary outcome is health-related (clinical, public health, or directly informing practice/policy).
The review type is a systematic review of intervention effects, prognosis, diagnostic accuracy, or aetiology — broadly the categories in our study type guide.
Data extraction has not yet begun.
You can wait 4–10+ weeks for confirmation without holding up the project.

Two practical notes that catch teams off-guard.

PROSPERO requires a defined primary outcome at registration. "Exploratory" or "outcomes to be determined by data" do not pass eligibility. If your review is genuinely exploratory in design, OSF is the more honest fit.

PROSPERO will reject reviews of methodology, reviews of reviews where the unit is reviews not studies, and most scoping/mapping reviews. The rejection is not a quality judgment — it is an out-of-scope determination. Read the eligibility page first, not the user guide.

Also: PROSPERO is not blockchain. Amendments are public, versioned, and timestamped. You can update the protocol after registration; what you cannot do is hide the history.

When to choose OSF

OSF Registries (Open Science Framework) is operated by the Center for Open Science. It runs as a self-service platform: a registration is created in minutes, with no eligibility gate.

OSF is the right choice when:

PROSPERO has rejected the review or it is plainly out of scope (scoping, mapping, methodology, qualitative-only, non-health).
You need registration immediately and cannot wait for PROSPERO's queue.
The review crosses disciplines — for example, a systematic review of educational interventions or a synthesis of policy research where the audience is not primarily clinical.
You want to register a non-traditional artifact alongside the protocol — e.g., a search algorithm, a labeled validation sample, an extraction template.

OSF supports several registration templates. For systematic reviews specifically, the OSF Registries' "Pre-Registration Template" with PRISMA-P fields is the convention. Some teams prefer the AsPredicted template for its brevity (8 questions); others use the longer OSF Standard Pre-Registration. All three produce a citable DOI and timestamped registration.

The trade-off is signal strength. OSF's lack of curation means a journal reviewer cannot infer "PROSPERO rejected this" from "registered on OSF" — but it also means OSF does not carry PROSPERO's specific stamp of clinical-review legitimacy. For non-clinical or out-of-scope reviews, this is irrelevant. For clinical reviews where PROSPERO would have accepted you, OSF is the weaker choice.

When to choose Cochrane Protocols

The Cochrane route is structurally different. A Cochrane protocol is not a registration — it is a peer-reviewed published document in the Cochrane Library. Registration is a byproduct.

The route is appropriate when:

The review is part of an accepted Cochrane Group's editorial program.
The team is willing to follow the Cochrane Handbook in full, including dual-reviewer screening, RoB 2 / ROBINS-I assessment (see our risk-of-bias guide), and Cochrane MECIR conduct standards.
The team can sustain the Cochrane editorial timeline (typically 6–18 months from protocol acceptance to first review draft).
The review benefits from the Cochrane brand on the published output.

The advantage is unambiguous: a Cochrane protocol is the strongest possible registration signal. The cost is that Cochrane reviews are slow, follow strict methodological constraints, and live inside an editorial process that is harder to navigate than either PROSPERO or OSF.

For most teams, Cochrane is not a choice between registries — it is a choice about which editorial pathway to take for the entire review. If the team is not committed to that pathway, PROSPERO + an open-access journal is the more flexible fit.

A decision tree that fits the actual constraints

What is the review type?
│
├─ Health-outcome systematic review (intervention, prognosis, DTA, aetiology)
│  ├─ Working with a Cochrane Group?
│  │  └─ Cochrane Protocols (peer-reviewed protocol pathway)
│  ├─ Need fast turnaround (< 2 weeks)?
│  │  └─ OSF (acknowledge weaker signal in journal cover letter)
│  └─ Standard timeline (4–10 weeks acceptable)
│     └─ PROSPERO
│
├─ Scoping or mapping review
│  └─ OSF (often paired with JBI scoping template)
│
├─ Methodology review (review of reviews, methods comparisons)
│  └─ OSF
│
└─ Qualitative synthesis or mixed-methods
   └─ OSF (PROSPERO accepts some qualitative SRs but eligibility varies)

What the registration entry should actually contain

All three registries align with PRISMA-P (the protocol-stage extension of PRISMA). The minimum viable registration covers:

Title and author team
Anticipated start and end dates
Funding source and conflicts of interest
Review question and PICOS or equivalent framework
Inclusion/exclusion criteria
Information sources (databases, grey literature, hand-search)
Search strategy (or commitment to publish it on completion)
Screening procedure (single vs dual; AI assistance if planned)
Risk-of-bias tool and rationale
Synthesis plan (narrative, meta-analysis methods, software)
Subgroup and sensitivity analyses planned

Two items are easy to under-specify and worth doing properly.

Search strategy. PROSPERO accepts a high-level summary at registration with the full strategy published later. The growing convention — pushed by PRESS peer review advocates and McGowan et al. (2016) — is to attach the full strategy at registration. It eliminates the post-hoc impression that the search was tuned to the result.

AI plan. Per the 2025 Cochrane position and RAISE, any planned AI-assisted task is a registerable item. Specify the tool, validation evidence, oversight, and override-rate red line. Reviewers are increasingly asking for this in the protocol entry, not the methods.

What changes between protocol and review

PROSPERO and OSF are versioned. Amendments are public. The healthy posture is to make changes openly with a brief justification.

The patterns reviewers flag as concerning:

Outcome switching. The primary outcome at registration is not the primary outcome at publication. This is the canonical selective-reporting failure.
Inclusion-criterion narrowing. Eligibility criteria added after screening began that exclude studies inconsistent with the conclusion.
Synthesis-plan change without rationale. Switching from narrative to meta-analysis (or vice versa) after seeing the data, with no explanation.

The cure is the same in every case: amend the registration, dated, with rationale, before the change is reflected in the review. PROSPERO and OSF both make this mechanically simple.

Common failure modes — and the fix

Three patterns we see repeatedly in protocols submitted for PRESS-style search peer review or AI-assistance gating before review kickoff.

"We'll register after we know what we're doing." This inverts the purpose of registration. Register when the protocol is locked, not when the review is finished. If you do not know what you are doing yet, you are still drafting the protocol — finish drafting first, then register.

"PROSPERO rejected us so we'll skip registration." OSF exists for exactly this case. The journal does not need PROSPERO; it needs a registration. Out-of-scope rejection is not the end of the road.

"We registered on multiple registries to be safe." Don't. It creates ambiguity about which entry is canonical and forces multiple amendment trails. Pick one. If you genuinely need cross-registry visibility (rare), reference the canonical entry in the others.

What this means for your next review

Three concrete steps before kickoff:

Decide registry on the basis of review type, not preference. Use the decision tree above; do not let "we always use X" make the call.
Lock the protocol before registering. Registration is the commitment; if you are still discussing inclusion criteria, you are not ready to commit.
Amend openly when needed. The amendment trail is the protection — silent changes are the failure mode.

Registration costs a few hours of work and weeks of patience. It is the cheapest defensibility insurance a systematic review can buy.