# How to Create an Exhibit List From Scanned Exhibits

- Author: Schema Law (https://www.schemalaw.com/)
- Published: 2024-06-30T07:39:32Z
- Updated: 2026-10-11T19:22:20Z
- Category: Litigation support
- Canonical: https://www.schemalaw.com/blog/how-to-create-an-exhibit-list

> How to create an exhibit list when the exhibits are scans: what courts ask for, which columns OCR text can fill, neutral descriptions and cross-references.

To create an exhibit list from scanned exhibits, start with the columns your court's rule or form asks for, then fill them from the documents themselves. Make each scan searchable, take the date, document type and Bates range from its first and last pages, write a short neutral description, and keep one master spreadsheet that every later version comes from.

The court forms quoted in this post have columns that are filled in at trial, and your own team finds exhibits by the same rows. This post covers how to build and maintain one when the exhibits began as paper. It explains document handling and is not legal advice. Court materials quoted here were read on each court's own site and can change.

## What does a court expect an exhibit list to contain?

An exhibit list, in each court document quoted in this section, contains a number for every exhibit and a short description that identifies it. Beyond that, the content is whatever your court's rule, your judge's order or the court's own form says. Find the form first, because its columns decide what you have to collect.

The federal starting point is brief. Among the pretrial disclosures in [Rule 26(a)(3)(A)(iii) of the Federal Rules of Civil Procedure](https://www.law.cornell.edu/rules/frcp/rule_26) is an identification of each exhibit, "separately identifying those items the party expects to offer and those it may offer if the need arises." The rule does not prescribe columns. Local rules and forms do.

Four documents show the range.

| Source | What it asks for |
| --- | --- |
| [Local Rule 3.07 of the Middle District of Florida](https://www.flmd.uscourts.gov/local-rules/rule-307-exhibits) | Under (b), "The list must sequentially list and briefly describe each exhibit." |
| The same court's [notice on pre-marking exhibits, with its exhibit list form](https://www.flmd.uscourts.gov/sites/flmd/files/judges/forms/flmd-kidd-exhibit-labels-and-list.pdf) | Columns for exhibit number, date identified, date admitted, witness and description |
| [Form AO 187, Exhibit and Witness List](https://www.uscourts.gov/sites/default/files/ao187.pdf) | Columns for plaintiff's number, defendant's number, date offered, marked, admitted, and a description of exhibits and witnesses |
| The District of Arizona's [Tucson instructions for exhibit lists](https://www.azd.uscourts.gov/sites/azd/files/judge-orders/TUC%20Exhibit%20Instructions.pdf) | "Exhibit descriptions should be brief, but descriptive enough to identify the exhibit." |

Read the instructions around a form as well as the form. The Middle District of Florida notice says "Exhibits must be consecutively numbered." The Tucson instructions say plaintiff's exhibits start at number 1 and defendant's exhibits "are numbered sequentially following the last number used by Plaintiff, or with a block of numbers assigned by the Courtroom Clerk, if applicable." AO 187 carries a footnote asking for "a notation as to the location of any exhibit not held with the case file or not available because of size." Each of those changes how you build the list.

State courts and hearings have their own versions. In California motion practice, for example, [rule 3.1110(f)(1) of the California Rules of Court](https://courts.ca.gov/cms/rules/index/three/rule3_1110) requires an index of exhibits that "must briefly describe the exhibit and identify the exhibit number or letter and page number."

### Two kinds of column

Sort the form's columns into two groups before you start.

- **Columns you fill before trial:** number, description, and anything your judge adds, such as Bates range, stipulations or objections.
- **Columns filled in the courtroom:** date identified, date offered, date admitted, witness. Leave them blank and leave room.

Only the first group can come from the documents.

## Which exhibit list columns can come from the scanned page?

The description, the document date, the document type, the author and recipient, the Bates range and the page count can all be read from a scanned exhibit once it has text. The exhibit number, objections, stipulations and anything about admission cannot. Those come from the trial team and the courtroom.

| Column | Comes from the page? | Where on the page |
| --- | --- | --- |
| Document date | Yes | Letterhead date, email header, signature block, form field |
| Document type | Yes | Title line, caption, form name |
| Author and recipient | Yes | Letterhead, "To" and "From" lines, signature block, address block |
| Title or subject | Yes | "Re:" line, subject line, heading |
| Bates begin and end | Yes, if stamped | Stamp on the first and last page |
| Page count | Yes | Count of pages in the file |
| Deposition exhibit number | Sometimes | Sticker on the first page |
| Trial exhibit number | No | Assigned by your side under the court's numbering rule |
| Expect to offer or may offer | No | Lawyer's decision |
| Objections and stipulations | No | Exchange with the other side |
| Offered, admitted, witness | No | Filled in during trial |

### Getting the page content into a form you can work with

A scan with no text gives you nothing to copy. Each exhibit has to be opened and retyped. Run OCR on the set first. How to convert a whole exhibit set and test it is covered in the guide to [building a searchable exhibit set](https://www.schemalaw.com/blog/ocr-trial-exhibits#building-a-searchable-exhibit-set).

With text behind the pages, the work for each exhibit is short:

1. Open the exhibit and its text side by side.
2. Copy the date, title and names from the first page into the row.
3. Go to the last page and read the Bates number and any signature date.
4. Check each copied value against the page image.

Schema Law reads an uploaded scan into text and tables that you can view page by page and export. It does not pick out the date or the author for you. You read them from the parsed first page and enter them. The same fields, and how to code them consistently across a large set, are the subject of the guide to indexing discovery documents, in the section on [objective coding of date, author, recipient and document type](https://www.schemalaw.com/blog/how-to-index-discovery-documents#objective-coding-date-author-recipient-and-document-type-from-the-page). If the case already has a discovery index, start the exhibit list from its rows.

### Where the page misleads

- **More than one date.** A letter can carry a typed date, a fax header date and a received stamp. Decide which one the list uses and apply it to every exhibit.
- **Stamps and stickers read as text.** OCR may read a Bates stamp as "ACME0O02101" with a letter O in place of a zero. Read the stamp on the image.
- **Handwriting.** Recognition is weaker on handwriting. A handwritten date needs your eyes.
- **Compilations.** An exhibit made of several documents has several dates and authors. Describe it as a set and give the range.

## Drafting exhibit descriptions from OCR text without characterizing the evidence

Describe what the document is, in the document's own terms: its type, its date, who it is from and to, and its title or subject line. Leave out what it proves. The converted text gives you the document's own words to build from, which keeps your adjectives out of the description.

Take a fictional case, Halvorsen v. Brightwater Marine Supply, about a late delivery of engine parts. Here is the same exhibit described three ways.

| Version | Description | Problem |
| --- | --- | --- |
| Characterizing | "Email in which Brightwater admits the shipment was late" | States a conclusion the other side disputes |
| Vague | "Email" | Does not identify the exhibit among forty emails |
| Neutral | "Email from D. Pruitt to K. Halvorsen, June 9, 2023, subject 'Order 4471 ship date'" | Identifies the document and nothing else |

The neutral version is assembled from four things on the page: the type, the "From" and "To" lines, the date and the subject line.

### A pattern for each document type

| Document type | Description pattern | Example |
| --- | --- | --- |
| Letter | Letter from [author] to [recipient], [date], re [subject line] | Letter from Brightwater Marine Supply to K. Halvorsen, May 2, 2023, re "Quote 2209" |
| Email or chain | Email (chain) from [author] to [recipient], [date or range], subject [subject] | Email chain between D. Pruitt and K. Halvorsen, June 9 to 12, 2023, subject "Order 4471 ship date" |
| Contract | [Title on the document], between [parties], dated [date] | "Supply Agreement" between Halvorsen Boatworks and Brightwater Marine Supply, dated January 15, 2023 |
| Invoice or record | [Type] no. [number], [issuer], [date] | Invoice no. 88213, Brightwater Marine Supply, June 20, 2023 |
| Photograph | Photograph, [neutral subject], [date if known] | Photograph of crated parts, undated |
| Compilation | [Type] ([count] documents), [date range] | Shipping records (12 documents), May to July 2023 |

### Rules of thumb for the wording

- **Quote titles and subject lines exactly**, in quotation marks, with their typos. They are the document's words.
- **Use names as they appear.** If the page says "D. Pruitt", do not expand it to a full name and job title you learned elsewhere.
- **No verbs of persuasion.** Admits, confirms, concedes, threatens, proves and fails all describe your argument.
- **No adjectives.** Late, defective, misleading and key belong in a brief.
- **Say "undated" or "unsigned"** when that is what the page shows. Do not supply a date from memory.
- **Keep it to one line.** The Tucson instructions ask for descriptions "brief, but descriptive enough to identify the exhibit," and a form's description column is narrow.

When you copy a subject line or title out of OCR text, compare it with the image. A description that misquotes the exhibit's own title is an error anyone holding the exhibit can see. Ask the lawyer responsible for the list to review the descriptions, because what counts as neutral in your case can be a legal judgment.

Add a column for the trial use of each exhibit. When you prepare the list for a pretrial conference, mark whether you expect to offer the exhibit or may offer it. If a motion in limine is planned, note that beside the exhibit number. When the parties sign a stipulation, mark the row and keep the scanned page with the list.

## Bates ranges, deposition exhibit numbers and trial numbers in one list

Give every exhibit one row keyed by its trial number, and put every other identifier on that row: beginning and ending Bates numbers, each deposition exhibit number it carried, and the page count. One row per trial exhibit means anyone holding any one of those numbers can find the other two.

A layout that works in a spreadsheet:

| Trial no. | Bates begin | Bates end | Pages | Also marked as | Description |
| --- | --- | --- | --- | --- | --- |
| 14 | BMS-000310 | BMS-000312 | 3 | Pruitt Dep. 6; Halvorsen Dep. 2 | Email chain between D. Pruitt and K. Halvorsen, June 9 to 12, 2023, subject "Order 4471 ship date" |
| 15 | BMS-000455 | BMS-000455 | 1 | Pruitt Dep. 9 | Invoice no. 88213, Brightwater Marine Supply, June 20, 2023 |
| 16 | none | none | 2 | none | Photographs of crated parts, undated |

### Checks that the cross-references make possible

- **Page count against Bates span.** Exhibit 14 runs from 310 to 312, so it should have three pages. A mismatch means a page is missing, a page was added, or the range is wrong.
- **Duplicates.** Sort by "Bates begin". Two trial numbers with the same starting Bates number are the same document listed twice, or two versions that need a note.
- **Overlaps.** A range that begins inside another exhibit's range is worth a look. One may be an excerpt of the other.
- **Unstamped exhibits.** Write "none" so a blank is never mistaken for an unfinished row.

### Where each number comes from

Read the Bates numbers from the first and last page images. If the stamp was recognized, a text search for the prefix finds it quickly, but confirm the digits by eye. How stamps are read, and when they are not, is covered in the guide on [how to OCR Bates numbered documents](https://www.schemalaw.com/blog/how-to-ocr-bates-numbered-documents).

Deposition exhibit numbers come from the sticker on the copy marked at the deposition and from any exhibit index in the transcript. Record the witness name with the number, since more than one deposition can have an Exhibit 6. Where two depositions used different copies of one document, say so in a notes column and decide which copy is the trial exhibit.

Trial numbers follow the court's scheme. Check whether your court wants one sequence for all parties, separate blocks, or prefixes, before you assign any.

## How do you keep the exhibit list in step with a changing exhibit set?

Keep one master list, give every exhibit a row that is never deleted, and tie each file name to its row. When an exhibit is added, withdrawn, renumbered or replaced, change the master first and regenerate everything else from it: the court form, the joint list, the binder tabs and the bookmarks.

### Rules for the master

1. **One file, one owner.** Everyone else gets exports. Edits come back to the owner.
2. **Never reuse a number.** A withdrawn exhibit keeps its row, marked "withdrawn" with the date. Reusing the number makes old notes and transcripts point to the wrong document.
3. **Keep a "former number" column.** If the court or a stipulation forces renumbering, the old number stays searchable.
4. **Version the scan as well.** If a better copy replaces a poor scan, note the date and the reason. The page count and Bates range may change with it.
5. **Date every export.** Put the date in the file name and in the footer.
6. **Log changes.** A second sheet with date, exhibit, what changed and who asked is enough.

### A reconciliation pass before each exchange or filing

| Check | How |
| --- | --- |
| Every row has a file | Sort the exhibit folder by name and compare it with the list, top to bottom |
| Every file has a row | Look for files in the folder with no matching number |
| Page counts match | Compare the "Pages" column with each file's page count |
| Descriptions still fit | Open any exhibit replaced since the last pass and reread its row |
| Bookmarks match | In a combined exhibit PDF, compare bookmark titles with the list's numbers and descriptions |

On that last check: if the exhibits are filed or lodged as a combined PDF, its bookmark titles and the list should use the same numbers and words. Generate the titles from the list's columns. Which courts require bookmarks and how to label them is covered in the guide on [how to bookmark a PDF for court filing](https://www.schemalaw.com/blog/how-to-bookmark-a-pdf-for-court-filing).

### Late additions

An exhibit added the week before trial can arrive as an unsearchable scan with no row. Give it the same intake as the rest: convert it, read its first and last pages, assign the next unused number, write the neutral description, and add a line to the change log.

## Exporting an exhibit list to the court's form or a joint list

Treat the court's form and the joint list as outputs of the master, never as the place you edit. Map the master's columns to the form's columns, export only those, leave the courtroom columns blank, and check the exported copy against the master row by row before it leaves the office.

### Mapping columns to a form

Using the Middle District of Florida form as the example:

| Form column | Master column | Filled when |
| --- | --- | --- |
| Exhibit No. | Trial no. | Before trial |
| Date Identified | none | In court |
| Date Admitted | none | In court |
| Witness | none, or sponsoring witness if your judge asks | In court or before |
| Description | Description | Before trial |

Bates ranges, deposition numbers and notes stay in the master unless your judge's order asks for them. If the order does, add the Bates range to the end of the description or as an extra column, whichever the order describes.

### Getting it into the court's file format

- **A fillable PDF form** such as AO 187 takes typed entries row by row. Paste from the spreadsheet one column at a time and check the row alignment after each paste.
- **A word-processing form** takes a pasted table. Paste as plain text into the form's existing table so its formatting holds.
- **A spreadsheet the judge provides** is the simplest. Copy values only.

Read the delivery instructions too. The Middle District of Florida rule calls for three copies to the judge and one to each opposing party before trial. The Tucson instructions call for an original and two copies to the courtroom clerk on the morning of trial, "Unless otherwise directed by the Court." Your court's numbers will differ.

### Merging into a joint list

When the court wants one list from all parties, agree on the columns and the numbering before anyone merges. Then:

1. Each side exports the agreed columns from its own master.
2. Combine them in one sheet with a column for the offering party.
3. Sort by Bates begin to find documents both sides listed.
4. For each duplicate, agree which number survives and record the other in "former number".
5. Add columns for objections or stipulations if the order asks for them.
6. Send the merged sheet back to each side to reconcile against its master.

If some of your exhibits are scanned schedules or ledgers and the list needs figures from them, the tables can be exported from the scan to a spreadsheet. In Schema Law, every plan includes every export format, Excel and CSV among them, and the [export formats on the homepage](https://www.schemalaw.com/#formats) show the full set. Whatever tool produced the text, the list is only as good as the check against the page image, so finish by opening ten rows at random and reading each against its exhibit.

## Try it on a real document

Upload a scanned PDF and export it as Word, a searchable PDF or plain text on the free plan. Your documents are never used to train an AI model. [Start free](https://www.schemalaw.com/api/auth/google)