Figure and Table Placement Is a Written Rule, Not a Guess.
Waraq · July 27, 2026 · RULE OF THE WEEK
In a PLOS ONE submission, each table sits inside the manuscript file directly after the paragraph where it is first cited, and every figure caption sits in that same spot, while the figure images themselves upload as separate high-resolution files. You do not collect the figures and tables at the end of the document, and you do not paste figure images into the running text. The two-column layout you see in the published article is built later by production staff. At submission your only job is to place each object where the journal's written rule says it goes.
Is the submission file the same as the published article?
No. The file you upload is a working draft for editors and reviewers, not the typeset page a reader downloads months later.
A published PLOS ONE article runs in two columns with figures dropped into whatever spot the page design allows. That layout is the output of production, and you never build it yourself. The manuscript you submit is double-spaced, single-column, and carries continuous line numbers, because it exists so reviewers can read and annotate it. Since the two documents serve different readers, they follow different placement conventions. Confusing one for the other is where a lot of first submissions go wrong. Authors either format the manuscript to imitate the printed journal, or they push every figure onto the last three pages because that is how the finished PDF looked.
What do the PLOS ONE placement rules actually say?
Placement is written down, and the two rules are short. For figures, PLOS ONE states: "Figure captions are inserted immediately after the first paragraph in which the figure is cited." For tables: "Tables are inserted immediately after the first paragraph in which they are cited."
Read those two sentences side by side, because they do not say the same thing. The figure rule moves the caption, not the image. The image uploads on its own, covered below, and the caption is the piece that lives in the text, right after the paragraph that first mentions the figure. The table rule moves the whole table into the text at the same trigger point, the first paragraph that cites it. In both cases the anchor is identical, the first citation in reading order, and the placement is "immediately after" that paragraph, not two pages down and not at the end.
The rule only works if you cite the object. You cannot place a table you never mention: with no callout in the text, there is no paragraph for it to sit behind. Every figure and table needs a numbered mention in the prose first.
Do figure legends go with the figure or in a separate list?
They go in the text, with the manuscript, not in a separate legends document and not inside the image file. The caption follows the paragraph that first cites the figure, exactly as the placement rule describes.
This is the part that trips people who have submitted elsewhere. Some journals ask for a stand-alone list of figure legends at the end. PLOS ONE does not. The caption stays in the running text so a reviewer reading top to bottom meets the figure's number, its title, and its explanation in one place. Two things stay behind when the image moves out to a separate file: the callout and the caption. The callout is the in-text mention, the plain "Fig 1" or "(Table 2)" that points the reader to the object. The caption is the titled description that sits after the paragraph. Keep both in the manuscript. Only the image travels separately.
Figure numbering has to line up with that citation order, the same discipline you apply to references. If you have wrestled with numbered citation order before, the logic behind Vancouver referencing is the same idea applied to the reference list.
Why does the journal want figures uploaded as separate files?
Because production needs the real image at print resolution, and a picture pasted into a text document cannot supply that. PLOS ONE asks for figures as separate TIFF or EPS files at 300 to 600 ppi, each named to match its citation, so "Fig 1" points to a file called Fig1.tif.
Tables are different because a table is text and rules, not a raster image. It stays in the manuscript file where an editor can copy a value or a reviewer can question a number. A figure is pixels, and pixels lose detail the moment they are squeezed into a Word file. Keeping the image as a separate high-resolution file protects the one thing production cannot recreate.
If separate uploads make you worry the reviewer sees a scattered pile of files, the submission system answers that. When you finish uploading, PLOS builds a single PDF for approval that merges the manuscript with low-resolution previews of each figure after the reference list, plus a link to download the high-resolution version. The proof reassembles everything so you can confirm it reads correctly before you approve. Separate files are how you upload. The PDF is how it gets read.
Before any of that, check the manuscript against the journal's written rules while you can still fix things quietly, rather than finding out at the desk-check stage that a formatting problem was visible from the start. A read against one specific journal's requirements is the kind of check Waraq runs on a manuscript before submission.
A placement checklist for the final document before upload
Work through the manuscript once with placement as the only question:
- Every figure and every table has a numbered callout in the text before it appears.
- Each table sits in the manuscript file immediately after the paragraph that first cites it.
- Each figure caption sits immediately after the paragraph that first cites the figure, with the image removed from the text.
- Figure files are separate TIFF or EPS uploads, named to match the citation, so Fig 1 becomes Fig1.tif.
- Captions live in the manuscript, not in a separate legends list and not inside the image.
- Numbering runs in citation order, with no gaps and no figure introduced out of sequence.
Placement is not a matter of taste, and it is not something the copyeditor quietly fixes for you. It is a written rule with a fixed anchor, and every journal states its own version. PLOS ONE wants the object right after its first citation. The reason to read the rule for the journal you are targeting is that the same care applies to abstract length, reference style, and every other line of the guidelines.