PLOS ONE Formatting Requirements 2026: Figures, Fonts, Files

PLOS ONE figures must be TIFF or EPS, 789–2250 pixels wide at 300–600 dpi.

The trade is that when the specification does bind, it is precise: a pixel range rather than a column width, a font whitelist, a resolution ceiling as well as a floor, and a set of TIFF flags that quietly fail if you get them wrong. This page collects all of it in the order you need it while preparing files.

PLOS ONE figure specifications at a glance

Everything in that table is stated on PLOS's own figure and submission pages. The rest of this page explains what each row means when you are actually building the file.

How wide can a PLOS ONE figure be?

The height maximum is the case where "the figure occupies the whole page and excludes the caption" — so a full-page figure is allowed, and the caption is not counted against it.

Three details decide whether your file lands cleanly inside that range:

PLOS does not prescribe a width, so pick one and commit before you lay out any labels: 13.2 cm for figures meant to read alongside the text, 19.05 cm for anything with multiple panels or dense data. Rebuilding a figure at a different width means re-laying-out every label, which is the expensive part.

What font and size can you use inside a figure?

This is a hard whitelist, and it catches people who have never had to think about it. PLOS states: "Use only Arial, Times, or Symbol font in 8-12 point."

Three fonts. Not "a clear sans-serif". Not "Helvetica or similar". If your plotting library defaults to DejaVu Sans, Computer Modern, Calibri or Helvetica — and most of them do — the text in your figures is out of specification and you will not notice, because it renders perfectly on your screen.

Two rules travel with it:

• Point size 8 to 12. Both ends bind: 7 pt annotations are as far out of specification as a 16 pt panel heading.
• Nothing about the paper goes inside the figure. "Do not include author names, article title, or figure number/title/caption within figure files." That information belongs in the caption, in the manuscript.

For EPS files, embed the fonts or convert the text to outlines — PLOS asks for this explicitly to prevent missing or badly rendered text, and notes that files produced by software such as MATLAB are best opened in Illustrator or Inkscape and converted before submission. Once text is converted to outlines it can no longer be edited, so convert on an export copy, never on your master.

The two things PLOS does not specify

Both come up constantly in figure preparation, and PLOS's figure page is silent on both. Here is what to do about each.

Which file formats and resolutions does PLOS ONE accept?

Two format traps are named explicitly:

• LaTeX-generated EPS is rejected. "PLOS does not accept vector EPS figures generated in LaTeX. Submit TIFF or EPS figures created in standard software." The route PLOS suggests is to compile to PDF, open the PDF in a graphics program, crop the figure and export as TIFF.
• RStudio's TIFF export is 72 dpi. Exporting a plot directly from RStudio to TIFF produces a 72 dpi file regardless of what you intended. Export EPS instead, or convert the figure afterwards.

On resolution, both ends have consequences — which is unusual, and the ceiling is the half people miss:

"Submit figures at the desired dimensions with a resolution no greater than 300-600 dpi."

• Below 300 dpi: "Resolution below 300 results in blurred, jagged or pixelated published figures."
• Above 600 dpi: "Resolution above 600 dpi may lead to resizing of the published figure." Submitting at 1,200 dpi to be safe is not safe; it can cost you control of the final size.
• Dense multi-panel figures: "If your figures have numerous pictures, charts, or small text, they will render best at a resolution of 600 dpi."
• The whole file inherits the worst element in it: "The quality of your figures is only as good as the lowest-resolution element present. If you created a 72 dpi line graph and placed it in a 300 dpi TIFF, the graph will look blurred, jagged, or pixilated."

That last one is the most common way a figure fails while looking fine. When assembling multi-panel figures in PowerPoint, Impress or Keynote, PLOS is specific about the mechanism: "Use an Insert tool to place figures. Do not drag/drop or copy/paste images into the file, because this results in a 72 dpi image."

Does PLOS ONE require a graphical abstract?

Four rules apply if you do upload one:

• It "must be derived from a figure or supporting information file from the submission" — a crop of an existing figure, or the whole figure. You cannot make a new image for the purpose.
• It should ideally be "high resolution, eye-catching, single panel."
• It should ideally avoid "added details such as text, scale bars, and arrows."
• "If no striking image is uploaded, we will designate a figure from the submission as the striking image."

How many figures can a PLOS ONE paper have?

This is the journal's single strongest differentiator on formatting, so it is worth being precise about what it does and does not mean.

Supporting Information figures

The size rule is the one worth reading twice. For main figures, 10 MB is a hard ceiling. For Supporting Information it is a recommendation, made "for ease of reader access".

Word and page limits

There is no word limit at PLOS ONE. What limits exist are structural, and most of them are character counts rather than word counts:

Four of these trip people up more than the rest:

A few layout rules complete the set: manuscript files must be DOC, DOCX or RTF and must not be locked or protected; LaTeX manuscripts are submitted as PDF; text is double-spaced and must not be in multiple columns; footnotes are not permitted; page and line numbers are required, numbered continuously.

Templates: LaTeX or Word?

The two paths are not symmetrical, and it is worth knowing which one you are on before you start.

• At submission: upload a PDF containing manuscript text, figure legends, tables and references. Figures do not go in the PDF — they are uploaded as separate files.
• After acceptance: you supply the LaTeX source, combined into a single .tex file.

Formatting quirks specific to PLOS ONE

Five things here that you will not find at other journals, in rough order of how much time they save or cost.

"PLOS One waives all formatting requirements until your manuscript has received a provisional Editorial Accept decision."

And then, in the same breath, what happens afterwards: "In order to proceed to publication, your figures must meet the requirements on this page. The more closely your figures adhere to these specifications, the more quickly your manuscript can be published once accepted."

• Flattened, no layers — and PLOS pre-empts the obvious mistake: "Figures with a single layer named 'layer 1' or 'layer 0' are in fact layered."
• No alpha channels.
• LZW compression is required — in GIMP, use Export rather than Save As and select LZW; in Photoshop, select LZW compression and "Discard Layers, and Save a Copy."
• Single page — "We cannot publish TIFF figures that span multiple pages."

• Authors "must provide the original, uncropped and minimally adjusted images supporting all blot and gel results."
• These go in a single PDF named S1_raw_images, uploaded as Supporting Information or deposited in a public repository with the DOI given in the data availability statement.
• That PDF has a 20 MB maximum — a different ceiling from the 10 MB main-figure limit, on the same page, governing a different thing.
• "Figure panels should include some background area above and below bands." And a companion rule: "A figure panel should not include composite images of bands originating from different blots, exposures, or gels."
• Non-compliance has a stated consequence: "the submission will be placed on hold while this is resolved."

Figure problems that trigger a revision request at PLOS ONE

Drawn from the requirements above, the failures that cost the most time between editorial accept and publication:

1. Text in the wrong font. Anything outside Arial, Times or Symbol, which means almost every plotting-library default. Fixing it usually means regenerating the plot, not editing the file.
2. A 72 dpi element inside a 300 dpi file. Drag-and-dropped panels, RStudio TIFF exports, and screenshots. The file passes a glance and fails the page.
3. A layered or alpha-channelled TIFF. Invisible locally, rejected at production.
4. A figure wider than 2,250 pixels, or one whose usable content is narrower than 789 once white space is excluded. Both are measurement errors rather than design errors.
5. Blots without the S1_raw_images PDF. This one does not produce a revision request — it puts the submission on hold.
6. Captions baked into the image. The caption belongs in the manuscript; a figure carrying its own title has to be re-exported.

Can you use AI to make a PLOS ONE figure, and what must you declare?

That policy is short enough to quote in the part that matters:

"Contributions by artificial intelligence (AI) tools and technologies to a study or to an article's contents must be clearly reported in a dedicated section of the Methods, or in the Acknowledgements section for article types lacking a Methods section. This section should include the name(s) of any tools used, a description of how the authors used the tool(s) and evaluated the validity of the tool's outputs, and a clear statement of which aspects of the study, article contents, data, or supporting files were affected/generated by AI tool usage."

Three things in that sentence decide what you actually write.

The hard line is separate and absolute: "The use of AI tools and technologies to fabricate or otherwise misrepresent primary research data is unacceptable." A schematic of a mechanism is not primary research data. A generated image standing in for a micrograph, a gel, or a plot of results is — and no disclosure makes that acceptable.

The stated consequence for getting this wrong is not a formatting query. Noncompliance "will be considered misrepresentation of methods, contributions, and/or results," and PLOS may notify the author's institution and reject, retract, or publish an editorial notice.

What this means for using SciFig

Two boundaries worth stating plainly:

• You still have to check it, element by element. A mislabelled receptor or a reversed arrow is your error at review and your correction after publication. The tool speeds up drawing; it does not transfer responsibility, and PLOS's wording asks you to describe the checking you did.
• Never for primary data. Charts and plots belong to your analysis pipeline, and micrographs belong to your microscope. Use figure tools for the conceptual layer — pathways, experimental design, mechanisms — and keep evidential imagery entirely outside them.

None of this reaches the figures you make outside a manuscript — grant applications, conference posters, slide decks, theses, lab websites. A journal's AI policy governs what that journal publishes; it says nothing about the rest of your year, which for most researchers is the larger share of the figures they make.

Build a PLOS ONE-ready figure

The whole specification, as a working sequence:

1. Decide the width first — 13.2 cm to align with the text column, 19.05 cm (2,250 px) for full page width. Everything else follows from this.
2. Set all figure text in Arial or Times at 8–12 pt. Check this before you have thirty labels to redo.
3. Export at 300 dpi, or 600 dpi if the figure is dense with panels or small text. Never above 600.
4. Confirm the file: TIFF or EPS, under 10 MB, RGB or grayscale, flattened, no alpha channel, LZW compressed, single page.
5. Name it to match the caption: Fig1.tif, Fig2.eps. Panels get lettered labels — (A), (B) — and are cited in text as Fig 1A, Fig 1B.
6. Put the caption in the manuscript, with a title of 15 words or fewer and any credits in the last sentence of the legend.
7. Run it through NAAS — main figures only, not Supporting Information.

None of this is required at first submission. All of it is required before publication, and the gap between the two is short.

Where to go next

• "Is PLOS ONE worth submitting to?" — an impact factor reported as 2.8, a Q2 placement in Multidisciplinary Sciences, and publication criteria that do not include novelty. PLOS ONE impact factor.
• "When exactly do these rules start binding?" — at the provisional editorial accept, which is the first of two accept decisions. PLOS ONE submission guide covers the sequence, the fees and what each submission status means.
• "How do other journals compare?" — Nature Communications requires vector line art at fixed 88 mm and 180 mm column widths and caps display items at 10, which is a very different regime. Nature Communications formatting requirements.

Frequently asked questions

Sources

Every specification here was read from PLOS's own author pages for PLOS ONE. Reviewed 2026.

• PLOS ONE — Figures — file formats, dimensions, resolution, colour mode, TIFF requirements, captions, NAAS, image integrity, blot and gel reporting, maps
• PLOS ONE — Submission Guidelines — length, title and abstract limits, manuscript file formats, layout rules, cover letter, Striking Image
• PLOS ONE — Supporting Information — file types, size guidance, naming and caption placement
• PLOS ONE — LaTeX — template, Overleaf, Vancouver style, submission and post-acceptance file requirements
• PLOS ONE — Editorial and Peer Review Process — the two-level accept decision that sets when formatting requirements bind