Co-Author Agreement: Key Elements & Sample

A co-author agreement is a document collaborators put together before substantial writing begins, setting out shared expectations for how a manuscript will be produced and credited. It is a different tool from a CRediT author statement, which is written during or after manuscript preparation to record, role by role, what each person actually did. A co-author agreement looks forward — anticipated order, expected contributions, who will correspond with the journal, a rough timeline, and what happens if something changes or a disagreement arises. A CRediT statement looks backward, documenting what was actually done once the work is finished. Well-run collaborations typically use both: an early agreement to set expectations, and a CRediT statement at submission to record what happened in practice.

This guide covers the elements a co-author agreement should address, why drafting one early materially reduces the risk of a later dispute, and a clearly-labeled illustrative example showing how those elements come together in practice.

Why draft an agreement before writing starts

CASRAI’s guide to resolving authorship order disputes covers this from the other direction — what to do once a dispute has already surfaced — and its own analysis of why these disputes happen points to the same root cause repeatedly: no early agreement. Author order is often left implicit until a manuscript is nearly finished, by which point positions feel fixed and renegotiation feels like an accusation. Long projects with personnel turnover, multi-site collaborations with different institutional ordering conventions, and student–supervisor power imbalances all make this worse the longer the question stays undecided.

A co-author agreement addresses this directly by moving the conversation to a point in the project where nobody yet has a stake in a specific outcome. It does not need to be a formal contract or require legal review in most academic settings — an email thread, shared document, or lab-standard template that every collaborator can see and has actively agreed to is generally enough to prevent the majority of disputes, per the same guidance. What matters is that it happens early, in writing, and with everyone’s actual agreement rather than one person’s assumption.

Neither the ICMJE authorship criteria nor the CRediT taxonomy (ANSI/NISO Z39.104-2022) require a written co-author agreement, and neither body publishes a mandatory template. Both are compatible with using one: ICMJE is explicit that authorship order is “to be decided collectively by the author group and not by editors,” and a documented early agreement is simply a practical way of exercising that collective decision deliberately instead of by default.

Key elements to include

1. Anticipated authorship order, and what could change it

State the order the group expects at the outset, and be explicit that it is anticipated, not final. More usefully, name the specific conditions that would trigger a renegotiation: a collaborator’s role expanding substantially beyond what was planned, a collaborator leaving the project before it’s finished, or a new contributor joining partway through. Committing to when the order gets revisited removes the ambiguity that turns an ordinary scope change into a dispute later.

2. Expected contributions per collaborator

Describe, in plain language or by mapping to the 14 CRediT contributor roles, what each person is expected to do: who designs the study, who collects or analyzes data, who drafts sections of the manuscript, who supervises. This does not need to be the final CRediT statement — it is a working plan, revisited as the project evolves, that later becomes the basis for the actual statement once the work is done. Naming roles early is also what gives the order-change trigger in element 1 something concrete to measure against.

3. Corresponding-author designation

Name who will serve as corresponding author and confirm they understand what the role involves: acting as the point of contact with the journal, obtaining co-author sign-off before submission and after any requested revisions, and, per ICMJE guidance, being responsible for including all appropriate co-authors and excluding inappropriate ones on submission. Decide this early rather than defaulting to whoever happens to submit the manuscript — the corresponding author role carries real ongoing responsibility, including after publication.

4. Timeline expectations

A rough shared timeline — target dates for drafting, internal circulation, and intended submission — keeps contributions on track and surfaces early if a collaborator’s availability has changed enough to affect their role. This does not need precision project-management detail; the goal is a shared reference point everyone can point back to if the project starts drifting.

5. Dispute-resolution process

Name a concrete first step for what happens if collaborators disagree: typically, revisiting the documented contributions together as a group before anyone treats the disagreement as unresolvable. If the group cannot reach agreement internally, both COPE and ICMJE are explicit that the next step is referral to the authors’ home institution(s), not the journal — editors are advised not to adjudicate authorship disputes themselves. Naming this path in the agreement up front means nobody has to figure out the escalation process for the first time while already in a dispute. See Resolving Authorship Order Disputes for the full mechanics of that escalation.

Illustrative example

The following is an illustrative composite, written to show how the five elements above come together in a single short document. It is not a real agreement from any specific lab, project, or institution, and the names are placeholders — adapt the structure and content to your own project and institutional norms rather than reusing the wording verbatim.

CO-AUTHOR AGREEMENT (draft, project kickoff)
Project: Soil microbiome response to reduced tillage – 3-site field study
Date: [kickoff date]

1. Anticipated author order
   1. R. Alvarez (lead: study design, field sampling)
   2. T. Okafor (lead: sequencing and bioinformatics analysis)
   3. M. Bergström (site access, local sampling coordination)
   4. J. Whitfield (senior author: supervision, funding)
   This order will be revisited if a collaborator's actual role changes
   substantially, if a collaborator leaves the project, or if a new
   contributor joins with a role comparable to an existing author's.

2. Expected contributions
   - Alvarez: Conceptualization, Investigation, Writing - original draft
   - Okafor: Formal analysis, Software, Writing - original draft (methods)
   - Bergström: Investigation (site 2), Resources
   - Whitfield: Supervision, Funding acquisition, Writing - review & editing
   Roles will be finalized as a CRediT statement at submission, reflecting
   what was actually done rather than what was planned here.

3. Corresponding author
   T. Okafor will serve as corresponding author: point of contact with the
   journal, responsible for circulating drafts and revisions to all authors
   for review and sign-off before each submission or resubmission.

4. Timeline
   Draft data collection complete: [date]. Internal draft circulated: [date].
   Target submission: [date]. Authors will flag as early as possible if
   their availability changes enough to affect this timeline.

5. If a disagreement arises
   Any author-order or credit disagreement will first be discussed directly
   among all co-authors, using the documented contributions above as the
   shared reference point. If not resolved internally, the matter will be
   referred to [home institution's research integrity office / department
   chair], consistent with COPE and ICMJE guidance that journals do not
   adjudicate authorship disputes.

Agreed by: [signatures or documented email sign-off from all listed authors]

A real agreement can be far shorter than this — even a short email restating these five points and confirmed by “agreed” replies from every collaborator satisfies the same purpose. The value is in committing to the five elements early and in writing, not in the format.

Revisiting the agreement

Treat the agreement as a living document, not a one-time form signed at kickoff and forgotten. The most useful points to revisit it are after a major role change, when a contributor joins or leaves, at any preregistration milestone, and again just before submission, when it’s worth confirming the final author list and order still match what was agreed and that everyone’s actual CRediT contributions are ready to be documented. A short check-in at each of these points is normally enough — the agreement doesn’t need to be renegotiated from scratch each time, only confirmed or updated.

How this relates to a CRediT statement

A co-author agreement and a CRediT statement serve different points in the same project’s timeline and should not be confused with each other. The agreement is a forward-looking plan, made before or during the work, about order, roles, and process. The CRediT statement is a backward-looking record, made at submission, of what each author actually did using the 14 standardized roles defined in ANSI/NISO Z39.104-2022. Neither replaces the other: an agreement with no eventual CRediT statement leaves the published record thin on contribution detail, and a CRediT statement with no prior agreement is exactly the late-stage, high-pressure situation most likely to produce a dispute. Using both, in sequence, is the practical shape most collaborations end up needing.

Frequently asked questions

Is a co-author agreement legally binding?

Not typically, and it generally doesn’t need to be. In most academic settings a co-author agreement functions as a professional and collegial commitment among collaborators, not a legal contract, and does not require legal review. Some institutional or industry-sponsored collaborations do formalize authorship terms within a broader research or data-sharing agreement that does carry legal weight — if your project involves that kind of formal agreement, check whether authorship terms are already covered there before drafting a separate document.

When should a co-author agreement be drafted?

At project kickoff, before substantial writing begins — the earlier the better, since the whole point is to reach agreement before anyone has a stake in a specific outcome. Revisiting it at major milestones (a role change, a contributor joining or leaving, preregistration, and just before submission) is more useful than treating it as a one-time document.

Does a co-author agreement determine the final CRediT statement?

No. It sets expectations going in, but the CRediT statement should reflect what each person actually did by the time the manuscript is submitted, even if that differs somewhat from what was originally planned. If a collaborator’s actual role changed substantially from the agreement, that’s exactly the kind of change the order-and-role sections of the agreement should already have a process for addressing.

Who should be included in the agreement — only people who will be authors?

Generally, everyone the group currently expects to meet ICMJE’s authorship criteria by the end of the project. Contributors who are expected to be acknowledged rather than credited as authors don’t need to be party to the agreement itself, though it’s good practice to be clear with them early about which category their involvement falls into, since that expectation can also shift as a project develops.

What if a collaborator won’t agree to the proposed order?

That’s what the dispute-resolution element of the agreement is for, applied early rather than after submission: revisit the documented contributions together as a group first. If it isn’t a one-off objection but a real, unresolved disagreement, treat it the same way as any other authorship dispute — see Resolving Authorship Order Disputes for the institutional-referral process both COPE and ICMJE point to when authors can’t reach agreement on their own.

Related CASRAI resources

Topic cluster

CRediT & Authorship Attribution — the full cluster hub for CRediT taxonomy, roles, and authorship-attribution content.

Co-Author Agreement: Key Elements and a Sample