Journal Article Highlights: Writing Guide

“Highlights” are a short, publisher-required set of bullet points — typically three to five, each capped at a strict character count — that authors submit alongside a manuscript to summarize its most novel findings for indexing, search, and rapid scanning. They are a distinct manuscript element from the abstract, the graphical abstract, and a video abstract: Highlights are plain text, not a paragraph, and not visual or multimedia. This guide covers what Highlights are, how their format requirements differ across publishers (verified directly against Elsevier’s own author guidance, since exact limits are frequently misstated online), how they differ from the other summary elements a manuscript may require, and practical guidance for writing them well.

What Highlights Are, and Why Journals Require Them

Highlights originated as an Elsevier/ScienceDirect convention and remain most strongly associated with that publisher, though the underlying idea — a short, structured, keyword-rich summary distinct from the abstract — now appears under various names across other publishers too. Per Elsevier’s own author guidance, Highlights are described as the “elevator pitch” of an article: a small set of bullet points that capture the novel results and, where relevant, new methods used in the study. They are submitted as a separate short file alongside the manuscript (not embedded in the manuscript text itself) and are typically requested at the revision or final-files stage rather than at initial submission.

Two purposes drive the requirement:

Discoverability. Highlights are indexed and displayed on the article’s landing page (e.g., on ScienceDirect) separately from the abstract, and their concise, keyword-forward phrasing is meant to surface in search results and database listings.
Rapid scanning. A reader deciding whether to invest time in the full abstract — an editor triaging submissions, a researcher scanning a table of contents, a database curator — can take in three to five short bullets faster than a 150-250 word paragraph.

Elsevier is explicit that Highlights are not used in the editorial/peer-review evaluation of the manuscript — they are a reader- and discoverability-facing element, produced after the science itself has already been judged on its merits.

Highlights vs. the Abstract vs. the Graphical Abstract vs. the Video Abstract

These four elements are frequently confused because journals often request several of them together, but each serves a different function and has a different form:

The abstract is continuous prose, generally 150-300 words, that summarizes the paper’s background, methods, results, and conclusions comprehensively enough to stand alone. Some fields formalize this into labelled sections — see CASRAI’s structured abstract format guide for how ICMJE, CONSORT, and PRISMA handle that. An abstract aims to be complete.
Highlights are plain-text bullet points, deliberately selective rather than comprehensive. Elsevier’s own guidance is explicit on this point: “Don’t try to capture all ideas, concepts or conclusions” — the goal is to surface only the most novel result(s), not to compress the whole abstract into list form.
The graphical abstract is a single image (a figure, diagram, or schematic) that conveys the paper’s central finding or method visually rather than in words. CASRAI’s separate graphical abstract guide covers design methodology and per-publisher image specifications — it is a visual-design task, not a writing task, and uses different software and skills than drafting Highlights.
The video abstract is a short recorded video, typically a couple of minutes, in which an author narrates the paper’s significance — production and presentation skills, not text-editing.

A single manuscript may be asked to provide any combination of these. They are not substitutes for one another, and drafting one well does not make the others unnecessary — each is evaluated and displayed separately on the published article page.

Format Requirements: Elsevier’s Own Specification (Verified)

Exact bullet-count and character limits are commonly misstated in secondary sources, so this section is drawn directly from Elsevier’s own current author guidance (elsevier.com/researcher/author/tools-and-resources/highlights) rather than from memory or a summary site:

Number of bullet points: three to five for most Elsevier journals; three to four specifically for Cell Press titles.
Character limit: each individual Highlight bullet may be no more than 85 characters, including spaces.
File handling: Highlights are uploaded as a separate short document, not typed into the manuscript body, and are collected at the revision/final-files stage rather than initial submission for most journals.
Not part of peer review: Highlights play no role in the editorial evaluation of the manuscript; they exist purely for the published, indexed article.

Because the 85-character limit includes spaces, it is tighter than it first appears — roughly the length of a single well-compressed sentence fragment, not a full sentence. Authors should count characters directly (most word processors show a live character count) rather than estimating.

Format Requirements Vary by Publisher — Always Check the Target Journal

Elsevier’s specification above should not be treated as a universal standard. Other publishers and journals use structurally similar but numerically different conventions, and some do not use Highlights at all:

Cell Press (an Elsevier imprint) tightens the count to three to four bullets while keeping the same 85-character limit — effectively a stricter subset of the general Elsevier rule, not a separate scheme.
AGU (American Geophysical Union) journals use an analogous but differently-specified element called “Key Points”: typically one to three statements, with a longer per-statement character allowance than Elsevier’s Highlights.
Some other publishers and journals request an equivalent element under a different name entirely — “Key Points,” “Author Summary,” or “Plain Language Summary” are common alternatives — each with its own bullet count, length limit, and audience framing (some are aimed at specialists, others explicitly at a lay/policy audience).
Many journals do not require any such element — Highlights are common in Elsevier’s STEM and life-science titles but are far from universal across scholarly publishing generally.

The reliable source is never a secondary summary (including this one) but the specific target journal’s current “Guide for Authors” page, searched for “Highlights,” “Key Points,” “Author Summary,” or “Plain Language Summary” — publishers revise these specifications, and a limit that was accurate for a given journal a year ago may no longer be current.

How to Write Effective Highlights

Lead with the single most novel or important finding. Because these bullets are often displayed in order and scanned quickly, put the strongest, most distinctive result first rather than building up to it the way an abstract’s narrative might.
Do not restate the title. A Highlight that just rephrases the paper’s title wastes one of only three to five available bullets and adds no new information for a reader who has already seen the title.
One idea per bullet. Resist compound sentences joining two findings with “and” — if a result needs its own bullet to be understood, give it one, within the total bullet-count limit.
Write for search and scanning, not narrative flow. Unlike abstract sentences, Highlights do not need to connect to one another grammatically. Favor keyword-forward phrasing a reader or search engine would actually query, over stylistic transitions.
Quantify where the character limit allows. A specific number (“reduced error rate by 34%”) is more informative than a vague claim (“substantially improved accuracy”) and often no longer in character count.
Avoid jargon, uncommon abbreviations, and acronyms that a reader outside the immediate subfield would need to look up — Elsevier’s own guidance flags this specifically, since Highlights are meant to be understandable at a glance.
Draft last, after the paper (and abstract) are finished. Highlights work best distilled from a completed, stable set of results and conclusions rather than drafted early and left unrevised as the paper’s framing evolves.
Count characters directly, including spaces, against the actual limit for the target journal — do not assume Elsevier’s 85-character figure applies elsewhere (see the previous section).

Illustrative Example

Illustrative example only — not a real published paper, study, or author. This is a composite created to demonstrate the writing pattern described above, not a case study of any actual research.

A weak Highlight, restating the title without adding information:

“A study of soil microbiome response to drought conditions”

A stronger Highlight for the same illustrative study, leading with the specific, quantified finding:

“Drought reduced soil microbial diversity by 28% within two weeks”

The second version fits comfortably within an 85-character budget, states an actual result rather than a topic label, and includes a number a reader or search index can act on.

Submission Checklist

Confirm the target journal actually requires Highlights, and under what name, by checking its current Guide for Authors — do not assume Elsevier’s rule applies elsewhere.
Confirm the exact bullet count (commonly three to five, three to four for Cell Press titles) and per-bullet character limit for that specific journal.
Count characters including spaces for every bullet, not just estimate them.
Check each bullet states a specific finding or method, not a restatement of the title or a topic label.
Check the bullets do not simply repeat the abstract’s sentences verbatim — they should be a distinct, selective distillation, not a duplicate.
Save Highlights as the separate file format the submission system requests (commonly a short, self-contained document), not embedded inside the main manuscript file.
Re-check length and content after any late-stage revision to the paper’s results or conclusions — Highlights drafted early can go stale.

Frequently Asked Questions

Are Highlights the same as keywords?

No. Keywords are typically single words or short phrases (e.g., 4-8 terms) used for indexing and database classification. Highlights are complete, if terse, statements — each a short sentence-length claim about a specific finding, not an isolated term.

Do Highlights count toward the manuscript’s word limit?

No. Highlights are submitted as a separate file/element from the manuscript text and are not counted against the article’s main word or page limit. They are still subject to their own, separate length limit (the per-bullet character cap).

Can Highlights be written before the paper is finished?

They can, but it is generally better practice to draft them last, once results and conclusions are stable — Highlights drafted early risk describing findings that shift during revision or peer review.

Do all journals require Highlights?

No. They are common in many Elsevier/ScienceDirect journals (and required as a distinct, more restrictive variant for Cell Press titles) but far from universal across scholarly publishing. Some publishers use an analogous but differently-named and differently-specified element (e.g., AGU’s “Key Points”), and many journals request no such element at all. Always check the specific target journal’s current author guidance.

Should Highlights repeat sentences from the abstract?

No. Highlights should be a distinct, selective distillation of the most novel finding(s), not a copy-paste of abstract sentences. Elsevier’s own guidance explicitly discourages trying to capture every idea from the paper in the Highlights — that is the abstract’s job.

See also CASRAI’s guides to the graphical abstract and the video abstract, and the structured abstract format guide, for the other manuscript-summary elements journals commonly request alongside or instead of Highlights.