A research protocol is the written plan for a study, set down before the work begins, that specifies what a research team intends to investigate and exactly how they will do it. It is the document reviewers, funders, and collaborators actually evaluate before any data collection starts, and it is the reference point everyone returns to afterward to check whether the study was conducted as planned. The most widely used formal definition comes from the ICH E6 Good Clinical Practice guideline, which defines a protocol as “a document that describes the objective(s), design, methodology, statistical considerations, and organization of a trial” (ICH E6(R2), Section 1.44) — a definition written for clinical trials but close in substance to how the term is used across most quantitative and qualitative research.
The term covers a wide range of documents in practice: a clinical trial protocol reviewed by an Institutional Review Board and a sponsor, a systematic review protocol registered with PROSPERO, a lab’s standard procedure for running an assay, or a short methods plan a graduate student writes before starting a thesis project. What unifies them is the same basic function — committing to a design and a method in advance, in enough detail that someone else could evaluate it, replicate it, or check the finished study against it.
What a research protocol actually contains
Protocol content varies by field and by how formal the review process is, but most protocols cover the same core elements, roughly in this order:
- Background and rationale — why the question matters and what existing evidence motivates the study.
- Objectives and hypotheses — stated specifically enough that a reader could tell in advance which results would support or contradict them.
- Study design and methodology — the overall design (randomized trial, cohort, qualitative interview study, systematic review, etc.), setting, and procedures.
- Participants or materials — eligibility criteria and sampling plan for human/animal subjects, or the materials and instruments involved in a non-human-subjects study.
- Data collection methods — instruments, measures, and the procedure for gathering data.
- Statistical or analytical plan — how the data will be analyzed, including how key decisions (exclusions, missing data, primary vs. secondary outcomes) will be handled.
- Ethical considerations — consent process, risk/benefit assessment, data protection, and any special protections for vulnerable groups.
- Timeline, roles, and dissemination plan — who does what, on what schedule, and how results will be reported.
Clinical trials have the most formalized version of this list: the SPIRIT 2013 Statement (Standard Protocol Items: Recommendations for Interventional Trials) sets out a 33-item checklist of minimum content for an interventional trial protocol, and is the reference point funders, IRBs, and journals commonly point authors to when specifying what a protocol submission must contain. A parallel checklist, PRISMA-P, does the same job for systematic review and meta-analysis protocols.
Why protocols get written
A written protocol serves several distinct purposes at once, which is part of why the document tends to be taken seriously even outside heavily regulated fields:
- Ethical and regulatory review. An IRB or Research Ethics Committee cannot assess participant risk, consent procedures, or data protection without a concrete plan to review — the protocol is what gets submitted for that review, not the eventual manuscript.
- Funder and sponsor accountability. Grant applications and clinical trial sponsorship agreements are built around a protocol; funders track whether the funded work matches what was proposed.
- Scientific integrity. Committing to hypotheses and an analysis plan before seeing the data is the mechanism that separates confirmatory testing from exploratory analysis dressed up after the fact — a distinction covered in more depth in CASRAI’s guide to preregistration of a study protocol. Preregistration is not a separate kind of document from a protocol; it is the practice of publicly time-stamping a protocol (or the hypothesis- and analysis-relevant parts of one) before data collection, specifically to make that commitment verifiable.
- Reproducibility. A sufficiently detailed protocol lets another team attempt the same study, and lets readers of the eventual publication evaluate whether the methods actually support the conclusions drawn.
Protocol vs. related documents
The word “protocol” gets used loosely enough that it’s worth distinguishing it from a few adjacent documents it’s often confused with:
| Document | What it actually is |
|---|---|
| Research protocol | The study plan itself — objectives, design, methods, analysis plan — written before the work starts. |
| Preregistration | The act of publicly time-stamping a protocol’s hypotheses and analysis plan before data collection, usually on a registry. |
| Grant proposal | A funding pitch built around a protocol, but also including budget, timeline, and significance/impact framing aimed at reviewers deciding whether to pay for the work. |
| Data management plan (DMP) | Covers how data will be organized, stored, and shared during and after the study — a required companion to a protocol for most funders, not a substitute for one. |
| Standard operating procedure (SOP) | A lab or facility’s reusable procedure for a recurring task (an assay, an instrument calibration), not tied to one specific study’s objectives. |
| Manuscript / final paper | Written after the study is complete, reporting what was actually done and found — ideally checkable against the protocol for any undisclosed deviation. |
Registering and publishing a research protocol
Once a protocol is written, there are several distinct things a research team might do with it, and they are not interchangeable:
- Preregistering it on a registry such as OSF Registries, AsPredicted, or ClinicalTrials.gov — this time-stamps the plan as evidence of what was decided in advance. See the full preregistration guide for how this works and which registry fits which study type.
- Registering a systematic review protocol with PROSPERO, the international registry for systematic review protocols, which reduces duplicate effort across teams and lets readers check a completed review against its planned scope.
- Publishing the protocol as a citable document on a platform such as protocols.io. This is a different act from registration: protocols.io assigns a DOI to a version of a step-by-step method so it can be cited and reused, but it does not itself function as a pre-data-collection timestamping registry the way OSF Registries or ClinicalTrials.gov do, and it does not peer-review what’s deposited.
- Publishing a standalone protocol paper in a journal that accepts them (BMJ Open, Trials, and similar venues run dedicated protocol-paper tracks), which puts the plan through peer review before the study is complete.
A single study can do more than one of these — for example, preregistering the analysis plan on OSF while also depositing the step-by-step laboratory method on protocols.io. They address different problems: preregistration guards against post-hoc reinterpretation of results, while a platform like protocols.io makes a method precisely reusable by other labs.
Who reviews a protocol
Depending on the study, a protocol can pass through more than one layer of review before work begins. Ethical review by an IRB or Research Ethics Committee evaluates participant protection — consent, risk/benefit, data privacy — and is the review most researchers encounter directly (CASRAI’s guide to the IRB/REC approval process covers what that review actually checks). This is a separate process from scientific-merit review, which some institutions run independently: certain NCI-designated cancer centers, for instance, require an institutional Scientific Review Committee to assess a protocol’s design, feasibility, and prioritization against competing open studies before it goes to the IRB at all — a requirement tied to their Cancer Center Support Grant funding, not a general rule across all research. Funders and sponsors typically review the protocol as well, for both scientific merit and budget/feasibility, and journals accepting protocol papers add editorial peer review on top of that.
Protocol amendments and deviations
A protocol is not expected to be followed with zero deviation from first draft to final report — it is expected that any departure from the plan is disclosed rather than quietly absorbed into the write-up as if it had been the plan all along. Formal amendments (a documented, usually IRB-approved change to the protocol itself, made prospectively) are distinct from deviations (an unplanned departure from the approved protocol, documented after the fact and assessed for whether it affected participant safety or data integrity). Both should be tracked and reported; a results section that never mentions any deviation from a study that ran for months is, in practice, unusual rather than reassuring.
Frequently asked questions
Is a research protocol the same thing as a preregistration?
No. A protocol is the plan itself. Preregistration is the act of publicly time-stamping some or all of that plan — typically the hypotheses and analysis plan — before data collection begins, on a registry that fixes the record so it can’t be edited retroactively without leaving a trace. Every preregistration is built from a protocol, but not every protocol is preregistered.
Do all studies need a written protocol?
Any study involving human or animal subjects reviewed by an IRB, IACUC, or ethics committee needs one, since that review has nothing else to evaluate. Funded research generally needs one because it’s the basis of the funding agreement. Purely exploratory or informal work may not have a formal protocol document, though most methodologists still recommend writing down the plan for anything intended for eventual publication.
What’s the difference between protocols.io and a preregistration registry?
protocols.io is built for sharing detailed, versioned, step-by-step methods so they’re citable and reusable by other labs — it doesn’t function as a pre-data-collection commitment registry and doesn’t peer-review what’s posted. A preregistration registry (OSF Registries, AsPredicted, ClinicalTrials.gov) exists specifically to lock a plan in place before data collection, as evidence the analysis wasn’t shaped by already knowing the results.
Can a protocol be changed after a study starts?
Yes, through a formal amendment process, usually requiring re-approval from whichever body approved the original protocol (IRB, sponsor, funder). What matters for research integrity is that the change and its rationale are documented and disclosed, not that the original plan is followed without exception.
Where should a systematic review protocol be registered?
PROSPERO is the standard international registry for systematic review protocols, and PRISMA-P is the reporting checklist commonly used to structure what goes into the protocol submitted there.
