# How to Make a Table of Authorities from a PDF

> Build a Table of Authorities when all you have is the final PDF: printed vs PDF page numbers, roman front matter, OCR pitfalls, and a manual method.

*Updated: 2026-09-24*

**The short answer:** you can build a Table of Authorities (TOA) from a PDF, but the page numbers must come from the *printed* page numbers on the brief, not from the PDF viewer's page counter. Search the PDF for every case, statute, rule and other authority, record the printed page of each appearance (including short forms, *Id.* and *supra*), group the entries by category, and format the list with dot leaders. By hand this takes hours on a long brief; a from-PDF tool does the extraction per page and leaves you to review. Either way, the PDF you index should be the exact PDF you will file.

This guide covers the situation Word's Mark Citation workflow does not: you have co-counsel's finished brief, an exported filing, or a scanned document run through OCR, and no marked-up Word file.

**Key facts: a TOA from a PDF**

- FRAP 28(a)(3) requires "a table of authorities—cases (alphabetically arranged), statutes, and other authorities—with references to the pages of the brief where they are cited." Local rules add their own details.
- The pages that count are the brief's printed page numbers. The PDF's page index usually differs because the cover and front matter come first.
- Front matter (tables, cover) is typically numbered in lowercase roman numerals; the body restarts at 1. That creates an offset you must subtract.
- Every appearance counts, not just the first full citation: short forms ("*Rehaif*, 588 U.S. at 230"), case-name-only references, *Id.* and *supra* all point to an authority on that page.
- Scanned briefs need a text layer (OCR) before any search or tool can read them, and OCR introduces character errors.
- *Passim* is used when an authority appears on so many pages that listing them is unhelpful; the threshold is a style or local-rule choice.

## Why PDF page numbers are the trap

A PDF viewer counts every sheet from the first: the cover is page 1, the table of contents is page 2, and so on. The brief itself prints different numbers at the foot of each page. A typical appellate brief has a cover with no number, front matter in roman numerals (i, ii, iii), and the body starting again at 1. If you copy the number from Preview's or Acrobat's page box, every reference in your TOA will be off by the length of the front matter. That is the single most common error in a hand-built TOA from a PDF.

**Example: printed page vs PDF page index**

| What the page is | Printed number on the page | PDF viewer page | Use in TOA? |
| --- | --- | --- | --- |
| Cover | none | 1 | No |
| Table of contents | i | 2 | No (not cited in) |
| Table of authorities | ii–iv | 3–5 | No |
| Statement of the case, first page | 1 | 6 | Yes: "1" |
| Argument page citing a case | 14 | 19 | Yes: "14", not "19" |
| Appendix page | A-3 (if numbered) | 41 | Only if your rules index the appendix |

The offset is not always constant. Inserted exhibits, a separately paginated appendix, or a blank page added during export can shift it partway through. Check the printed number on the page you are looking at, every time, or verify the offset at several points in the document.

## Three kinds of PDF, three levels of difficulty

| PDF type | Can you search it? | Main risk |
| --- | --- | --- |
| Exported from Word, Docs or Pages ("native") | Yes, reliably | Page offset; hyphenated line breaks splitting a case name |
| Scanned, then OCR'd | Mostly | Misread characters: "ld." for "Id.", "S" for "§", broken "v." |
| Scanned, no OCR (image only) | No | Nothing to search until you run OCR |

### OCR caveats

- **Section symbols** (§, §§) are often read as "S", "$" or "88". Search for the statute number as well as the symbol.
- **Id.** is often read as "ld." (lowercase L) or "Id,". Search both.
- **"v."** in case names can come through as "v" or "y."; search by party name, not by the full caption.
- **Italics and small caps are lost**, so you can no longer spot case names by typeface.
- **Footnotes** are read in a different order from the body text; make sure you record the page the footnote appears on.
- **Reporter volumes and pin cites** can gain or lose digits. Check any citation that looks odd against the page image.

## The manual method: building a TOA from a PDF in Preview or Acrobat

- **1. Confirm the offset.** Note the PDF page where printed page 1 falls, and check it again near the end of the brief.
- **2. List candidate authorities.** Skim the argument and write down every case, statute, regulation, rule and constitutional provision. If the brief already has a draft TOA or TOC, start from it, but do not trust its page numbers.
- **3. Search each authority.** In Preview, press ⌘F; results appear in the sidebar with page thumbnails. In Acrobat or Acrobat Reader, Shift-⌘F (Advanced Search) lists every hit with its page. Search by a distinctive party name, not the full citation, so short forms are caught.
- **4. Catch *Id.* and *supra*.** Search "Id." and "supra" separately and, for each hit, decide which authority it refers to by reading the preceding citation.
- **5. Record printed pages.** Write the printed page number of each hit, not the viewer page. Remove duplicates on the same page.
- **6. Categorize and alphabetize.** Cases, constitutional provisions, statutes, regulations, rules, other authorities, or whatever headings your court expects.
- **7. Decide on *passim*.** Apply your rule consistently.
- **8. Format.** Put page numbers at a right-aligned tab stop with a dot leader, then proofread the TOA against the brief once more.

The weak point of the manual method is step 3. Party-name searches find long citations but miss references such as "the *Rehaif* Court held" when you searched for the other party, and they miss *Id.* entirely. Plan a final read-through of the brief with the TOA beside it.

> **Common mistake:** "The page number in the PDF viewer is the page number for the TOA." It almost never is. Front matter numbered in roman numerals pushes the body pages later in the file. Always use the number printed on the page.

## What a from-PDF TOA tool has to do

Any tool that claims to build a TOA from a PDF, including a general chatbot, should be judged on the same checklist. If it cannot show you where each page reference came from, you have to check every one by hand anyway.

| Requirement | Why it matters |
| --- | --- |
| Per-page text extraction | Each authority must be tied to the page it sits on, not to the document as a whole. |
| Printed page mapping (roman front matter, body restart) | Otherwise every reference is shifted by the offset. |
| Short-form grouping | "*Greer*, 141 S. Ct. at 2097" and the full citation are one entry. |
| *Id.* and *supra* resolution | These appearances count, and they point to the last or named authority. |
| Categories | Cases, statutes, rules, regulations, constitutional provisions, other authorities. |
| *Passim* option | Long lists of pages for a frequently cited case need a consistent rule. |
| Review before export | Extraction is never perfect; you need to see and correct every entry. |
| No invented entries | The table must contain only authorities actually in the brief. |

## How BriefAuthority approaches it

BriefAuthority is built for exactly this case. It reads the final PDF page by page in your browser, maps each page to its printed number (including roman-numeral front matter), and detects cases, statutes, regulations, court rules, constitutional provisions and other authorities. A review screen then lists every authority with the pages it appears on, where you merge short forms, link *Id.* and *supra* references, re-categorize or exclude entries. It exports a formatted .docx TOA with category headings, dot leaders, page references and optional *passim*. The brief is processed on your computer and is not uploaded. *Specification as designed; confirmed at launch.*

In an internal pre-launch test on two U.S. Supreme Court merits briefs (*Greer v. United States*, No. 19-8709; *Bouarfa v. Mayorkas*, No. 23-583), the citation engine found **129 of the 131 cases** listed in the briefs' own professionally prepared Tables of Authorities. Page lists matched exactly for 77–84% of cases before review; the misses were mostly short references by case name alone (for example, "Rehaif"), which the review screen is built to catch. Statutes and rules have not yet been benchmarked.

What it does not do: it does not check whether a citation is good law or correctly cited, and it does not edit your brief. It needs a PDF with a text layer, so an image-only scan has to be OCR'd first. Formatting and indexing only — verify every citation.

> **Where BriefAuthority fits:** BriefAuthority is pre-launch and not available to buy yet. The planned price is $39 one-time or $9 per brief. If you want a TOA from your final PDF, on your Mac, without uploading the brief, join early access at /early-access to get the launch price first.

## Frequently asked

**Can I make a table of authorities from a PDF instead of a Word file?**

Yes. Search the PDF for each authority, record the printed page number of every appearance (including short forms, *Id.* and *supra*), group entries by category and format them with dot leaders. Tools that read the PDF page by page can automate the extraction, but you should still review every entry.

**Why are my table of authorities page numbers off by a few pages?**

You probably used the PDF viewer's page counter. The cover and roman-numeral front matter come before printed page 1, so the viewer's number is higher by the length of the front matter. Use the number printed at the foot of each page.

**How do I find every page a case is cited on in a PDF?**

In Preview use ⌘F; in Acrobat or Acrobat Reader use Advanced Search (Shift-⌘F), which lists all hits with pages. Search by a distinctive party name so short forms are found, then search "Id." and "supra" separately and assign each hit to the right authority.

**Does a scanned brief work for building a TOA?**

Only after OCR gives it a text layer. Expect errors: "ld." for *Id.*, "S" for "§", broken "v." and wrong digits in volumes and pin cites. Check suspicious citations against the page image.

**Do Id. and supra references go in the table of authorities?**

The page they appear on does. *Id.* and *supra* are appearances of the authority they refer to, so that page belongs in the authority's page list. Follow your court's rules and your office's style if they say otherwise.

**When should I use passim in a table of authorities?**

Where your court allows it, many offices switch to *passim* at five or more page references — that threshold comes from Microsoft Word's built-in "Use passim" option. The U.S. Supreme Court's Rule 34.2 says a table of authorities should not use passim (added in the 2023 rules and kept in the rules effective March 16, 2026), and the D.C., Tenth and Eleventh Circuits bar it too (D.C. Cir. R. 28(a)(2); 10th Cir. R. 28.3(C); 11th Cir. R. 28-1(d)). Follow your court's rule.

**Can ChatGPT build a table of authorities from my PDF?**

A general chatbot can produce a list, but it does not guarantee that page numbers come from the printed pages, that every appearance is included, or that no authority is invented, and you would be uploading the brief to a third party. Whatever you use, check each entry against the brief.

**Related guides**

- [How BriefAuthority works](/how-it-works) — The PDF-to-.docx pipeline and the review screen.
- [How to make a Table of Authorities on a Mac](/how-to-make-a-table-of-authorities-on-a-mac) — Options when Word for Mac does not officially support TOAs.
- [Table of Authorities in Google Docs or Pages](/table-of-authorities-google-docs-pages) — No built-in feature: what to do instead.
- [Table of Authorities without uploading the brief](/table-of-authorities-without-uploading) — For confidential briefs that cannot leave your computer.
- [Table of Authorities rules](/table-of-authorities-rules) — What FRAP 28 and the Supreme Court rules require.
- [Early access](/early-access) — Get the launch price first.
- [Id. and supra in a table of authorities](/id-supra-table-of-authorities)
- [Passim: which courts ban it](/passim-table-of-authorities)
- [Can ChatGPT make a table of authorities?](/chatgpt-table-of-authorities)

**Sources**

- [Federal Rule of Appellate Procedure 28 (Cornell LII)](https://www.law.cornell.edu/rules/frap/rule_28)
- [Supreme Court of the United States — Rules and Guidance](https://www.supremecourt.gov/filingandrules/rules_guidance.aspx)
- [Microsoft Support — Create a table of authorities](https://support.microsoft.com/en-us/word/create-a-table-of-authorities)
