Acceptance deserves a moment of relief. When the proofs arrive, a focused check of meaning, production queries and corrections can help the team finish the publication process carefully.
September 24, 2026 · 4 minutes
Take the relief, then change the task
An acceptance email can feel like permission to stop thinking about the paper for a while. We can enjoy that moment and still leave room for the work that follows.
When the proofs arrive, our suggestion is to give them a different kind of read. Instead of polishing the argument again, ask whether the version readers will see still communicates the accepted work accurately. Start with the proof instructions and deadline in the publisher's email, then agree who will check which parts.
This is a practical editorial workflow, not a claim that every journal uses the same production system or permits the same changes.
Read the instructions before making corrections
Elsevier's proof-correction guidance, updated August 14, 2026, directs authors to use the method specified in their proof email and states that only the corresponding author can submit corrections. Its instructions cover several systems rather than one universal interface. [1]
Before distributing tasks, identify the authorized submission route. We suggest keeping one shared correction list so the person returning the proof can distinguish agreed corrections from open questions. Ask colleagues to give a location, the current text and the requested replacement, rather than a general instruction to check a section.
A visible update date does not establish that these requirements were newly introduced on that date. The useful point is that the current instructions should guide the current proof.
Check meaning where small details matter
For a hypothetical clinical research paper, imagine that a table heading in the proof gives a measurement in milligrams while the accepted table gives it in micrograms. That is a specific discrepancy to resolve, not an opportunity to rewrite the Discussion.
Our proposed check starts with the accepted files and follows a few important details into the proof: units, numerical values, symbols, figure labels and the text that explains them. Check whether a superscript, footnote or table heading still points to the right information. Read the caption alongside the figure, not as a separate fragment.
We also suggest checking author names, affiliations and declarations in the displayed proof. These checks are editorial recommendations, not a substitute for the publisher's own instructions or the authors' scientific review.
Treat production queries as decisions
Elsevier's guidance describes answering copy-editor queries within its proofing systems. For Neptune specifically, unresolved queries prevent submission, and the guidance warns that authors cannot make further changes in Neptune after submitting corrections. Those details should not be generalized to every proofing platform. [1]
Our recommendation is to separate simple corrections from substantive questions. If a proposed edit affects the interpretation of a result, ask the relevant coauthor to review it and contact the journal or production team about how to proceed. Do not quietly turn an unresolved scientific issue into a typesetting correction.
Before submitting, read the assembled corrections once more. Check that two authors have not requested conflicting replacements and that the answer to a production query matches the correction made in the text.
Do not leave a known error for after publication
Nature Support explains that post-publication changes follow a formal correction process and may not always be possible. Its guidance asks authors reporting an error to identify the article, describe the problem and explain the requested correction. This is publisher-specific guidance, not a rule that all later changes are impossible. [2]
If a known issue remains unresolved at proof stage, our advice is to flag it through the designated contact before returning the proof. If an error is discovered later, use the journal's correction procedure rather than assuming the published file can simply be replaced.
A practical proof-review checklist
- Read the proof email, return deadline and correction instructions.
- Identify who will consolidate and submit the team's corrections.
- Compare important numbers, units and symbols with the accepted files.
- Read figures, captions, table headings and footnotes together.
- Check displayed author details and declarations.
- Resolve production queries or raise a clearly described question with the journal.
- Review the combined corrections for conflicting instructions.
- Retain the submitted correction record and any confirmation provided by the system.
We suggest treating this as a bounded final task, with enough attention to protect the meaning of the work and enough restraint to avoid reopening every sentence. The team has already brought the paper a long way. The proofs deserve a careful finish.
Sources
Put the context to work
JournalReady helps you review a manuscript against the submission you intend to make. It does not replace editorial judgment or predict acceptance.
Start manuscript review