Skip to main content
Manuscript Preparation10 min readUpdated Jul 17, 2026

How to Write Journal Highlights for a Manuscript

A source-backed workflow for writing manuscript highlights that meet journal rules, avoid generic claims, and make the paper's actual findings easy to scan.

By Erik Jia
Author contextFounder, ManusightsView profile

Readiness scan

Find out if this manuscript is ready to submit.

Run the Free Readiness Scan before you submit. Catch the issues editors reject on first read.

Check my manuscriptAnthropic Privacy Partner. Your manuscript is never used to train any model.See example reports
Working map

How to use this page well

These pages work best when they behave like tools, not essays. Use the quick structure first, then apply it to the exact journal and manuscript situation.

Question
What to do
Use this page for
Getting the structure, tone, and decision logic right before you send anything out.
Most important move
Make the reviewer-facing or editor-facing ask obvious early rather than burying it in prose.
Common mistake
Turning a practical page into a long explanation instead of a working template or checklist.
Next step
Use the page as a tool, then adjust it to the exact manuscript and journal situation.

Quick answer: If you need to know how to write journal highlights, make 3 to 5 short bullets only if your target journal asks for them, then check the exact journal limit before upload. Elsevier's common rule is 3 to 5 highlights, each no more than 85 characters including spaces. Cell Press commonly uses 3 to 4. AGU uses Key Points instead, with 1 to 3 statements and a 140-character limit. The best highlights state the paper's real findings, method advance, or practical consequence without repeating the title or overstating the Results.

Use this guide with the journal submission keyword workflow, submission readiness checklist, and manuscript quality check. If your highlights are hard to write, the problem is often not the bullet format. It is that the abstract, figures, and claims do not yet agree.

Use this guide when you are preparing the final manuscript package before you submit. For a manuscript-specific check, run a highlights and claim-calibration review before you upload the file.

From our manuscript review practice

Highlights should state what the paper found, not what the paper is about.

Who this is for

This is for authors preparing a manuscript package for journals that request Highlights, Key Points, or similar short finding statements. It is not a universal formatting rule. Some journals require highlights at submission, some encourage them, some use another name, and some do not use them at all.

The reader job is practical: finish the upload package without creating administrative friction or claim overreach. A highlight is usually not where the editor decides the paper's fate, but it can expose whether the manuscript's story is specific, supported, and easy to scan.

First, check which highlight system the journal uses

Do not apply an 85-character Elsevier rule to every journal. Publisher systems differ.

Publisher or journal system
What the current source says
What to do before upload
Elsevier Highlights
Elsevier describes Highlights as 3 to 5 bullets, each 85 characters or fewer including spaces
Prepare a separate editable Highlights file when the target Guide for Authors asks for it
Cell Press Highlights
Cell final-file guidance describes 3 to 4 bullets, each no more than 85 characters
Treat them as part of the Cell Press summary package with the graphical abstract and eTOC material
AGU Key Points
AGU allows 1 to 3 Key Point statements, each at most 140 characters, with no abbreviations
Write complete finding statements, not abbreviated Elsevier-style fragments
PLOS and many Nature Portfolio journals
The general instructions do not use Elsevier-style Highlights as a standard manuscript section
Do not add a Highlights file unless the specific journal asks for it

The safest rule is simple: open the target journal's current author instructions and search for "Highlights," "Key Points," "Author Summary," and "Plain Language Summary." These are different artifacts. Do not merge them into one generic bullet list.

The journal highlights workflow

Use this workflow after the Results and figures are stable.

Step
Action
Decision test
Risk if skipped
1. Extract the real findings
Pull 6 to 8 candidate statements from the Results, figure legends, and conclusion
Can each statement point to a figure, table, model, or analysis?
A highlight becomes a topic slogan rather than a result
2. Remove background
Delete bullets that only say why the field matters
Would the sentence still be true before this study existed?
The highlight wastes a slot on context
3. Match the journal format
Apply the exact count, character limit, and file-upload rule
Does the target journal allow this format today?
The upload package is returned or manually corrected
4. Make each bullet self-contained
Include the subject, action, and finding when space allows
Can a reader understand it without the title?
The bullet becomes cryptic outside the manuscript
5. Check claim strength
Compare every verb against the Methods and Results
Is "shows," "predicts," "associates," or "improves" the honest verb?
The highlight overclaims the paper
6. Align with title and abstract
Ensure the highlights support the same central claim
Do the title, abstract, highlights, and Figure 1 tell one story?
The package looks strategically unfocused

Write findings, not topic labels

A weak highlight names a subject. A useful highlight names a result.

Weak highlight
Better direction
Why
"New biomarker for disease detection"
Name the biomarker, disease context, and validated performance signal
Specificity makes the claim checkable
"Machine learning improves diagnosis"
Name the model task and comparison baseline
"Improves" needs a comparator
"Catalyst performance is enhanced"
Name the material, reaction, and measured improvement
The reader needs the system, not the adjective
"We study inflammation in patients"
Name the cohort, mechanism, and observed association
A study topic is not a finding
"A new method is proposed"
State what the method measures or enables
Novelty is not useful unless the function is visible

The character limit forces discipline, but it should not force vagueness. If the paper has several claims, choose the three to five that carry the manuscript's actual contribution. If no claim can fit cleanly, the abstract may also be too broad.

A practical template

Use this template as a drafting scaffold, then adapt it to the journal's current rules.

Highlights

• Single-cell profiling identifies IL-17A+ fibroblasts in inflamed tissue.
• Catalyst aging reduces hydrogen yield under humid operating conditions.
• The model improves lesion detection versus the radiologist-only baseline.
• The mechanism is supported by rescue assays and dose-response controls.

Before upload, replace these examples with your own findings, count characters including spaces, and check whether the journal wants a Word file, a text-box entry, or no highlights at all.

In our pre-submission review work on journal highlights

In our pre-submission review work, journal highlights are one of the fastest ways to detect a package-level claim problem. Manusights submission analysis reads them against the title, abstract, figures, tables, Methods, statistical analysis, cover letter, and target-journal instructions. We see the same specific failure pattern across fields: the highlight is short enough to pass a character counter but not honest enough to pass reviewer scrutiny. A good set of highlights makes the paper easier to classify. A weak set often exposes the same issue that will later appear as a reviewer objection.

Journal highlights that repeat the title. This is common when authors write highlights first. The title already carries the broad claim. Use highlights to show the strongest evidence path: population, method, mechanism, comparison, table, or validated result. If the title says the paper identifies a pathway, a useful highlight should say which experiment, cohort, figure, or control supports that pathway.

Journal highlights with stronger verbs than the Results support. "Demonstrates" is not the same as "is associated with." "Predicts" is not the same as "correlates with." "Restores" is not the same as "partially improves in one assay." Editors and reviewers notice when short summary artifacts inflate the manuscript's evidence. We check each verb against the Results section, sample size, confidence interval, control condition, and primary endpoint.

Journal highlights that list methods instead of findings. A bullet that says "RNA-seq and proteomics were performed" does not tell the reader what was learned. If the method is the contribution, state what the method enables, detects, or resolves. If the paper is not a methods paper, the highlight should usually connect the method to a result in a figure, table, model, or supplementary analysis.

Journal highlights that hide the negative or bounded result. A result can be useful even when conditional, null, or narrower than expected. Do not erase the boundary. A precise bounded finding is stronger than a broad unsupported one. If the Discussion admits that the effect appears only in one subgroup, the highlight should not imply a universal mechanism.

Journal highlights that conflict with the cover letter. This happens when the cover letter sells broad clinical, engineering, or policy significance while the highlights describe a narrow assay or modeling result. The mismatch makes the submission package feel strategically split. Align the cover letter, abstract, highlights, Figure 1, and conclusion around the same evidence-backed contribution before upload.

Journal highlights that ignore the target journal's format. A strong Elsevier-style Highlights file can still be wrong for AGU Key Points, a PLOS Author Summary, or a Nature-family journal that does not ask for Highlights. We check the target journal's current author instructions before evaluating the wording because the right artifact depends on the venue.

In practice, the useful question is not only "are these 85 characters or fewer?" The useful question is whether the highlight, abstract, Figure 1, statistical analysis, and cover letter make the same evidence-backed promise. That is the non-obvious analysis authors often miss when they treat Highlights as a formatting afterthought.

Check whether your title, abstract, highlights, and claims tell one supported story.

What to do in the next 30 minutes

  1. Open the target journal's author instructions and record the exact highlight artifact: name, count, character limit, upload location, and file type.
  2. Draft 6 to 8 candidate bullets from the Results, figures, tables, and conclusion.
  3. Delete any bullet that only gives background, repeats the title, or names a method without a finding.
  4. For each remaining bullet, write the figure, table, or analysis that supports it.
  5. Replace overstrong verbs with evidence-matched verbs.
  6. Count characters, including spaces and punctuation.
  7. Read the title, abstract, highlights, and Figure 1 together. If they point to different claims, fix the manuscript package before upload.

Think twice before finalizing highlights when

Do not treat highlights as cosmetic. Pause before upload when:

  • the highlights claim a mechanism that appears only in the Discussion;
  • the title promises one contribution and the highlights emphasize another;
  • one bullet depends on a secondary or exploratory analysis;
  • the journal uses Key Points, Author Summary, or Plain Language Summary instead of Highlights;
  • the bullets cannot fit the limit without losing scientific meaning.

In those cases, a submission readiness check is usually more useful than another round of copyediting. The issue is alignment across the submission package, not only word choice.

Readiness check

Run the scan to see how your manuscript scores on these criteria.

See score, top issues, and what to fix before you submit.

Check my manuscriptAnthropic Privacy Partner. Your manuscript is never used to train any model.See example reports

Evidence basis

This page was checked on July 17, 2026 against current public guidance from Elsevier, Elsevier Support, Cell Press final-file requirements, AGU publication instructions, PLOS ONE submission guidelines, Nature Scientific Reports submission guidelines, and Wiley manuscript-preparation guidance.

Official facts used here are narrow: Elsevier-style Highlights commonly use 3 to 5 bullets and an 85-character limit; Cell Press commonly uses 3 to 4 bullets and an 85-character limit; AGU Key Points use a different 1 to 3 statement format with a 140-character limit; and several publishers require authors to follow the specific target journal's instructions rather than a universal highlights rule.

Bottom line

Good journal highlights are not mini-abstracts. They are short, source-compliant statements of what the paper actually found. Write them after the evidence is stable, keep the journal's exact format in front of you, and make every bullet traceable to a result.

Frequently asked questions

Check the target journal's current instructions. Elsevier's general guidance usually asks for 3 to 5 highlights, while Cell Press commonly asks for 3 to 4. AGU uses Key Points rather than Elsevier-style Highlights and allows 1 to 3 statements.

For Elsevier-style Highlights and Cell Press Highlights, the common limit is 85 characters including spaces. Do not assume that limit for every publisher. AGU Key Points, for example, use a different 140-character limit.

Use highlights for the paper's core results, distinctive method, validated mechanism, or practical consequence. They should not restate the title, list background, advertise the topic, or make claims that the Results do not support.

Write a rough list after the Results are stable, then revise it after the abstract and title are final. The highlights should align with both without copying either one.

References

Sources

  1. Elsevier: Highlights of journal articles
  2. Elsevier Support: include Highlights with your manuscript
  3. Cell final file requirements
  4. AGU text and graphics requirements
  5. AGU submission checklists
  6. PLOS ONE submission guidelines
  7. Scientific Reports submission guidelines
  8. Wiley manuscript preparation guidelines

Final step

Find out if this manuscript is ready to submit.

Run the Free Readiness Scan. See score, top issues, and journal-fit signals before you submit.

Anthropic Privacy Partner. Your manuscript is never used to train any model.

Internal navigation

Where to go next