Most readers never notice a good book index, and that is exactly the point. A well-built index works silently, guiding a reader straight to the one passage they need without a wasted minute. A bad index is noticed immediately: entries that lead nowhere, names spelled three different ways, page numbers pointing at the wrong discussion, and an alphabetical order that follows no alphabet at all. For nonfiction authors, the index is often the difference between a book that gets used and a book that gathers dust. This guide shows you how to format a book index for print and ebook in 2026: what to index, how to structure entries, how to sort them, how to handle page numbers, and how print and ebook indexes differ.
By the end, you will know how to format a book index that professionals would be happy to sign: choosing entries, building a clean hierarchy, setting columns and indents in print, linking locators in an ebook, and proofing the whole thing before you upload your files.
Quick answer: To format a book index, list the book’s key topics, names, and concepts as alphabetized main entries with indented subentries, attach accurate page numbers (or linked locators in an ebook), sort letter by letter while ignoring articles and punctuation, and set the result in two narrow columns with hanging indents. Build the index from the final paginated file, never from a draft, and proof every locator against the text.
In this guide
- What a Book Index Is (and How It Differs from a Table of Contents and a Glossary)
- What to Index (and What to Leave Out)
- Entry and Subentry Structure
- Alphabetical Sorting Rules
- Page Numbers and Ranges
- Print Index Formatting: Columns, Indents, and Runover Lines
- Ebook Index Differences: Linked Locators Instead of Page Numbers
- Print vs Ebook Index at a Glance
- Tools and Software for Building an Index
- Common Mistakes When You Format a Book Index
- Your Book Index Proofing Checklist
- Frequently Asked Questions
- Conclusion: An Index Is a Promise to the Reader
What a Book Index Is (and How It Differs from a Table of Contents and a Glossary)
A book index is an alphabetized list of the topics, names, places, and concepts discussed in a book, each paired with the pages where that topic is discussed. It is a finding tool, not a summary: it answers the question “where does this book talk about X?” in seconds. It sits in the back matter, after the appendices, and it is one of the first things a nonfiction reader checks before buying.
The index is often confused with two other back-matter tools, and the confusion causes real formatting errors. A table of contents lists chapters in the order they appear; an index lists topics in alphabetical order, regardless of where they appear. A glossary defines terms; an index locates them. The table of contents shows structure, the glossary explains meaning, and the index finds things. Our guide to formatting a table of contents covers the chapter-list side of the book in detail, and it is worth reading alongside this guide when you lay out the back matter. Our front matter and back matter guide maps the full trade sequence the index sits within.
Why does an index matter so much? For reference books, textbooks, histories, biographies, and how-to guides, the index turns a one-time read into a working tool. Readers return to indexed books again and again, and reviewers notice when an index is missing or sloppy. A missing index in a serious nonfiction book reads as an unfinished book.
What to Index (and What to Leave Out)
The first decision when you format a book index is what earns an entry. The rule is simple: index topics a reader would plausibly look up, and only pages where the topic is genuinely discussed. A passing mention in a single sentence is not a discussion, and giving it a locator wastes the reader’s time. If a reader follows an entry and finds nothing useful, they stop trusting the index.
Index proper names, places, organizations, key concepts, and important terms. Index the main ideas of each chapter, not every word in it. For a history book, index people, battles, treaties, and movements. For a cookbook, index ingredients, techniques, and dish names. Think like your reader: what words would they type into a search box if this book were a website? Those are your entries.
Just as important is what to leave out. Do not index the author of the book, the book’s own title, or generic words like “introduction” or “chapter.” Do not index every occurrence of a word that appears on nearly every page; a term discussed throughout the book needs to be broken into subentries instead, so the reader can go straight to the facet they want. Do not index footnotes unless they carry substantive discussion, and never index the front matter, the table of contents, or the index itself.
Entry and Subentry Structure
The basic unit of an index is the main entry: a single word or phrase, followed by its locators. Under it sit subentries, indented, that break the topic into facets. Subentries can themselves carry sub-subentries, indented one level further, though a third level should be used sparingly. The hierarchy turns a flat list of words into a map of the book’s ideas.
Consider a history book’s entry for a general. The main entry carries the name. Beneath it, subentries name campaigns, battles, and controversies: “early career,” “Gettysburg campaign,” “relationship with Lincoln.” Each subentry carries only the pages where that facet is discussed, so a reader who wants one battle does not wade through every page where the general is mentioned.
Cross-references are the second structural tool. “See” sends the reader from a term you did not use to the term you did: “Automobiles. See Cars.” “See also” points from one entry to a related one. Use cross-references deliberately: every synonym a reader might try should either be an entry or a “see” reference. Main entries begin at the left margin in the same typeface as the body text, subentries indent one em space, and entries are set in sentence case except for proper names.
Alphabetical Sorting Rules
Sorting looks easy until you meet the edge cases, and edge cases are where indexes fail. The standard professional rule is letter-by-letter sorting: compare entries character by character, ignoring spaces and punctuation, so that “Newark” (newark) files before “New York” (newyork) because at the fourth letter, “a” comes before “y.” Word-by-word sorting treats each word as a unit and sorts “New York” before “Newark” because the first word “New” is shorter. Most trade publishers use letter-by-letter; whichever you choose, apply it consistently.
Articles at the start of an entry are ignored for sorting: “The Great War” files under “G,” not “T.” Initialisms file alphabetically as written, so “NATO” sits under “N.” Numbers at the start of an entry usually file before letters, in numeric order, or are spelled out and filed by the spelled form. Names with prefixes need a decision too: “de Gaulle” files under “G” in most American styles and under “D” in some European ones.
Page Numbers and Ranges
Page numbers are the locators of a print index, and their accuracy is non-negotiable. A single wrong page number breaks trust in the entire index. The iron rule: build the index only from the final, paginated file. Never index a draft, because pagination shifts with every edit, and never reuse an index from an earlier edition without checking every locator against the new pagination.
Ranges need a consistent style. The standard is to give the full span for the first number and abbreviate the second according to house style: “123-45” for pages 123 to 145, or “123-145” if your style spells ranges out in full. Either is acceptable; mixing them is not. Use an en dash, not a hyphen, for ranges. For non-consecutive pages, separate the locators with commas: “45, 67-69, 112.” Match the locator to the size of the discussion: do not give a range for a single passing mention, and do not give a single page for a discussion that spans three.
Print Index Formatting: Columns, Indents, and Runover Lines
A print index is almost always set in two columns. Two narrow columns suit short lines of text, let the reader scan quickly, and keep the index compact. The columns sit within the normal text measure, with a gutter between them, and the index continues the book’s running heads and page numbers like any other back-matter section.
Each entry uses a hanging indent: the first line sits at the left margin of the column, and any runover lines indent by one em. This is the single most recognizable feature of a printed index, and it is what makes scanning possible: the eye travels down the left edge, reading main entries, and dips into the indented subentries only when it finds the right one. Type size is usually one point smaller than the body text, in the same typeface family, and each new letter section begins with a bold letter heading set on its own line.
Ebook Index Differences: Linked Locators Instead of Page Numbers
An ebook has no fixed pages, so an ebook index cannot use page numbers. This is the fundamental difference, and it changes both the build and the reading experience. When you format a book index for ebook, each locator becomes a hyperlink that jumps to the exact passage in the reflowable text. The reader taps “Gettysburg campaign” and lands on the paragraph, not on a page. Done well, an ebook index is better than a print index.
Building a linked index requires care in the EPUB file. Each locator must point to a stable anchor in the text, and those anchors must survive conversion and validation. Do not fake it by linking every locator to the start of a chapter; a link that lands far from the discussion is worse than no link. Anchor at the paragraph level wherever your workflow allows it, and test every link in a real ereader before release. The index itself must also be reflowable: never embed the print index as an image in an EPUB.
Print vs Ebook Index at a Glance
The table below summarizes the key differences between a print index and an ebook index. Keep it beside you when you plan the back matter, because the two formats are built from the same entry list but finished very differently.
Tools and Software for Building an Index
You have three realistic options for building an index, and the right one depends on your budget and your patience. The first is dedicated indexing software, such as Cindex or SKY Index, which professional indexers use. These tools manage entries, subentries, cross-references, sorting, and locator formatting, and they export clean files for print and ebook production. They cost money and take learning, but for a long or complex book they pay for themselves.
The second option is to index inside your layout program. InDesign has a built-in indexing feature that lets you mark entries in the text and generate the index from those markers, with page numbers updating as pagination changes. Word has a simpler version of the same idea. These tools are adequate for short books with modest indexes, but they grow clumsy as entry counts climb. The third option is to hire a professional indexer.
Common Mistakes When You Format a Book Index
The most common index mistakes are easy to name and easy to avoid once you know them. Indexing from a draft instead of the final pagination, so locators point at the wrong pages. Indexing every mention instead of every discussion, so entries drown in useless numbers. Sorting inconsistently, so the same kind of entry files two ways. Forgetting cross-references, so readers who try a synonym find nothing. And in ebooks, linking locators to chapter starts instead of to the passage itself.
Two subtler mistakes deserve mention. The first is the orphaned subentry: a subentry whose main entry was deleted or renamed, leaving it stranded under the wrong heading. The second is the inconsistent name form: “Kennedy, John F.” in one entry and “John F. Kennedy” in another. Pick one form for every person, place, and concept, and use it everywhere.
Your Book Index Proofing Checklist
Run through this checklist before you export your files, for print and ebook alike:
- Build the index from the final paginated file, never a draft
- Every main entry has at least one locator
- Every subentry sits under the correct main entry
- Sorting is letter-by-letter throughout, with articles ignored
- Every cross-reference points to an entry that exists
- Page ranges use en dashes and one consistent abbreviation style
- Print columns use hanging indents and bold letter headings
- Every ebook locator links to the exact passage
- Spot-check twenty entries at random by following every locator
- One name form for every person, place, and concept
- No orphaned subentries from deleted or renamed main entries
- No page numbers in the ebook index
Proofing the index against the book is the step most self-published authors skip, and it catches the errors readers actually notice. If any locator you spot-check is wrong, do not assume the rest are right; check them all. An index is a promise, and a broken promise in the back matter colors the whole book. Pair a clean index with a properly formatted title page and the book opens and closes with equal authority.
Frequently Asked Questions
Conclusion: An Index Is a Promise to the Reader
A book index is a small part of a book that carries an outsized share of its reputation. Readers may never praise a good index, but they remember a bad one, and reviewers and librarians judge a nonfiction book partly by it. The way to format a book index is not complicated: choose entries a reader would look up, build a clean hierarchy, sort consistently, attach accurate locators, and proof the result against the files. Do that for print and ebook alike, and your book will repay the effort every time a reader finds what they want in seconds.
Get a book index readers can actually use
ArrowUpz formats complete back matter to trade standards — alphabetized indexes with clean entry and subentry hierarchies, consistent sorting, accurate print page numbers, and linked ebook locators, for both print and ebook editions.
Basic Formatting — $97 per book
Professional Formatting — $147 per book
Premium Formatting — custom quote




