Where do the citations actually live in the PDF?

In a few places depending on the style: footnotes at the bottom of each page, endnotes gathered at the end of a section/document, or a reference/bibliography list at the end, plus in-text citations that point to them. So extracting "footnote citations" means first locating where the references are (footnotes, endnotes, or the reference list) and then pulling them out. Footnotes are interleaved with the body text per page, which makes them a bit more work to isolate than a clean reference list. So step one is identifying the citation format the paper uses, which tells you where to extract from — the bottom of pages, an endnotes section, or the bibliography.

How do I extract the footnote/endnote text?

Pull the text from the note regions: for a born-digital PDF the text is real and extractable; for a scan, OCR it first. Footnotes sit at the bottom of pages (often in smaller type), so you extract those areas across pages; endnotes and reference lists are contiguous and easier. Then you have the raw citation text to work with. The complication is that footnotes are mixed with body text and page furniture, so extraction can interleave them — you may need to separate the note text from the main text. So extract the text (OCR scans), isolating the footnote/endnote/reference content, as the raw material for parsing into individual citations.

How do I turn the raw text into a usable citation list?

Parse it: separate the run of note text into individual citations and, ideally, capture the fields (author, title, year, source) for each into a structured list or spreadsheet. Citations follow style conventions (which helps parsing), but real footnotes are messy — they include "ibid.", "op. cit.", page numbers, and discursive content mixed with references — so parsing is imperfect and needs cleanup. The goal is a clean list of the actual references. Where citations include DOIs, those are the most reliable anchor. So parse the raw note text into separate, fielded citations, expecting to clean up the artifacts (ibids, commentary) that footnotes carry alongside the references.

Why must I verify the extracted citations?

Because extraction and parsing both make errors, and you are usually building a bibliography or doing citation analysis where wrong references matter. OCR (for scans) misreads, footnote text interleaves with body text, and parsing can split or merge citations or grab the wrong fields — so the output is a draft, not a finished bibliography. Verify each citation against the source: confirm authors, titles, years, and that "ibid."-type references were resolved to what they point to. A bibliography with mangled or wrong citations is a real scholarly error. So treat the extracted list as a strong starting point you check and correct, not a finished product — the tool gathers, you verify.

What about "ibid.", "op. cit.", and repeated references?

These are the tricky part of footnote extraction: scholarly footnotes use "ibid." (same as the previous), "op. cit." / short forms (refer to an earlier full citation), and other back-references, so the literal note text often does not contain the full citation — it points to one earlier. To build a complete reference list you have to resolve these back to the full citations they reference, which is a manual interpretation step the extraction cannot reliably do. So expect to chase down what each "ibid."/short-form refers to and substitute the full reference. This is exactly why footnote-citation extraction needs human verification — the abbreviated back-references are meaningful only in context.

How do I get the citations into my bibliography or reference manager?

Once you have a verified, fielded citation list, format it into your required style or import it into a reference manager — driving from DOIs where available gives the most reliable, complete bibliographic data. So the end of the workflow is: verified citation data → format into the style you need, or import to a reference manager that can pull full records (especially from DOIs). The extracted-and-verified list is the input; the formatter or manager produces the finished bibliography. This is the same as any citation workflow, with the front-end twist that you sourced the citations by extracting them from a PDF's footnotes rather than entering them by hand.

Is it safe to do this with unpublished papers online?

For unpublished or confidential papers, prefer a tool that processes files locally. ScoutMyTool extracts text, OCRs, extracts to a spreadsheet, and formats citations entirely in your browser tab, so the paper never leaves your machine. For pre-publication work, confirm the tool does not upload before using it — and always verify the extracted citations against the source.

6 min read

How to extract footnote citations from an academic PDF

By ScoutMyTool Editorial Team · Last updated: 2026-05-22

Pulling the citations out of an academic PDF’s footnotes into a usable list is genuinely handy — for a literature review, citation analysis, or building a bibliography — but footnotes are messier than they look. The references live at the bottom of pages (or in endnotes / a reference list), interleaved with body text, and laced with “ibid.” and short-form back-references that point elsewhere. This guide walks the realistic workflow: locating the notes, extracting the text (OCR scans), parsing the references, resolving the back-references, and — essentially — verifying each citation, because extraction and parsing both err. Then format the verified list into a bibliography or import it to a reference manager.

Locate → extract → parse → verify

Step	Detail
Locate the notes	Footnotes per page, or endnotes / reference list
Extract the text	Pull the note text (OCR if scanned)
Parse the references	Separate citations; capture author/title/year
Verify	Check each against the source; fix parse errors
Format / import	Bibliography or reference-manager import

Step by step — citations into a list

Identify the citation format. Footnotes per page, endnotes, or a reference list — this tells you where to extract from (see extracting footnotes).
OCR if scanned. Recover note text with PDF OCR; footnotes are small type, so verify carefully.
Extract the note text. Pull the footnote/endnote/reference content, separating it from body text and page furniture.
Parse into fielded citations. Separate individual references and capture author/title/year into a list with PDF to CSV (see extracting structured data).
Resolve back-references. Chase “ibid.”/“op. cit.” to the full citations they point to — a manual step.
Verify each citation. Check authors/titles/years against the source; the rigor of bulk citation extraction.
Format or import. Drive citations (ideally from DOIs) with the Citation Formatter or a reference manager — see citation export.

Extract footnotes: isolating the note content.
Bulk citation extraction: author/date across many papers.
Paper management & citations: building the bibliography.
Academic research workflow: the scholarly context.
Extract structured data: into a spreadsheet.
Citation Formatter: format the verified list in your browser.
All ScoutMyTool PDF tools: the full toolkit.

FAQ

Where do the citations actually live in the PDF?: In a few places depending on the style: footnotes at the bottom of each page, endnotes gathered at the end of a section/document, or a reference/bibliography list at the end, plus in-text citations that point to them. So extracting "footnote citations" means first locating where the references are (footnotes, endnotes, or the reference list) and then pulling them out. Footnotes are interleaved with the body text per page, which makes them a bit more work to isolate than a clean reference list. So step one is identifying the citation format the paper uses, which tells you where to extract from — the bottom of pages, an endnotes section, or the bibliography.
How do I extract the footnote/endnote text?: Pull the text from the note regions: for a born-digital PDF the text is real and extractable; for a scan, OCR it first. Footnotes sit at the bottom of pages (often in smaller type), so you extract those areas across pages; endnotes and reference lists are contiguous and easier. Then you have the raw citation text to work with. The complication is that footnotes are mixed with body text and page furniture, so extraction can interleave them — you may need to separate the note text from the main text. So extract the text (OCR scans), isolating the footnote/endnote/reference content, as the raw material for parsing into individual citations.
How do I turn the raw text into a usable citation list?: Parse it: separate the run of note text into individual citations and, ideally, capture the fields (author, title, year, source) for each into a structured list or spreadsheet. Citations follow style conventions (which helps parsing), but real footnotes are messy — they include "ibid.", "op. cit.", page numbers, and discursive content mixed with references — so parsing is imperfect and needs cleanup. The goal is a clean list of the actual references. Where citations include DOIs, those are the most reliable anchor. So parse the raw note text into separate, fielded citations, expecting to clean up the artifacts (ibids, commentary) that footnotes carry alongside the references.
Why must I verify the extracted citations?: Because extraction and parsing both make errors, and you are usually building a bibliography or doing citation analysis where wrong references matter. OCR (for scans) misreads, footnote text interleaves with body text, and parsing can split or merge citations or grab the wrong fields — so the output is a draft, not a finished bibliography. Verify each citation against the source: confirm authors, titles, years, and that "ibid."-type references were resolved to what they point to. A bibliography with mangled or wrong citations is a real scholarly error. So treat the extracted list as a strong starting point you check and correct, not a finished product — the tool gathers, you verify.
What about "ibid.", "op. cit.", and repeated references?: These are the tricky part of footnote extraction: scholarly footnotes use "ibid." (same as the previous), "op. cit." / short forms (refer to an earlier full citation), and other back-references, so the literal note text often does not contain the full citation — it points to one earlier. To build a complete reference list you have to resolve these back to the full citations they reference, which is a manual interpretation step the extraction cannot reliably do. So expect to chase down what each "ibid."/short-form refers to and substitute the full reference. This is exactly why footnote-citation extraction needs human verification — the abbreviated back-references are meaningful only in context.
How do I get the citations into my bibliography or reference manager?: Once you have a verified, fielded citation list, format it into your required style or import it into a reference manager — driving from DOIs where available gives the most reliable, complete bibliographic data. So the end of the workflow is: verified citation data → format into the style you need, or import to a reference manager that can pull full records (especially from DOIs). The extracted-and-verified list is the input; the formatter or manager produces the finished bibliography. This is the same as any citation workflow, with the front-end twist that you sourced the citations by extracting them from a PDF's footnotes rather than entering them by hand.
Is it safe to do this with unpublished papers online?: For unpublished or confidential papers, prefer a tool that processes files locally. ScoutMyTool extracts text, OCRs, extracts to a spreadsheet, and formats citations entirely in your browser tab, so the paper never leaves your machine. For pre-publication work, confirm the tool does not upload before using it — and always verify the extracted citations against the source.

Verify extracted citations. Footnote extraction and parsing are imperfect, and “ibid.”/short-form back-references must be resolved by hand. This article covers the extraction workflow; always verify each citation against the source.

Citations

Wikipedia — “Citation,” the references you are extracting. en.wikipedia.org/wiki/Citation
Wikipedia — “Note (typography),” footnotes and endnotes. en.wikipedia.org — Note (typography)
Wikipedia — “Bibliography,” the output you are building. en.wikipedia.org/wiki/Bibliography

Footnotes into a verified bibliography

Extract, parse, and format citations with ScoutMyTool’s in-browser tools — the paper never leaves your machine. Always verify each citation against the source.

Open the Citation Formatter →