How to edit text in a PDF

First, work out which kind of PDF you have

Open the file in any reader and try to drag-select a sentence. The result tells you which of the four methods below is even available to you:

• Individual words highlight. There is a real text layer. Every method here is on the table.
• Nothing highlights, or the whole page highlights as one block. The page is a picture — a scan, a phone photo, or an export that flattened everything. You need OCR before anything else.
• Text selects but pastes as gibberish. The embedded font has a broken or missing character map. Editors will fight this too; rebuilding usually costs less time than fixing it.

Method 1: edit in place with a PDF editor

This is what most people mean by editing a PDF: click into the text and change it, leaving the rest of the page alone. What you get depends on the tool.

Acrobat Pro

The only tool that reliably reconstructs paragraphs and reflows the surrounding lines when your replacement is longer than the original. It is a subscription, at roughly twenty dollars a month, which is a lot for one date change and nothing at all if you do this weekly.

LibreOffice Draw — free, desktop

Draw imports each PDF page as a canvas of text boxes you can click into. For swapping a name, a figure or a date it is genuinely good, and nothing leaves your machine. Longer edits expose the limitation: each box is laid out independently, so text that outgrows its box does not push the next paragraph down — it overlaps it. Exporting also re-encodes the whole file, so expect a different file size and, occasionally, different line breaks elsewhere on the page.

Browser-based editors

Fast and require no install, but check two things before you use one on anything private. First, most of them upload the file to a server to do the work. Second, the free tier is usually a couple of documents a day, and some stamp the result. Neither is a problem for a flyer; both are for a payslip.

The font trap, in every editor. A PDF normally embeds a subset of each font — only the glyphs the document actually used. Type a character that was never in that subset, and the editor has to substitute another font for it. The line comes out at a slightly different weight or width, and nothing warns you. Always compare the edited line against the one above it.

Method 2: annotate instead of editing

If the change is really “add a note”, “sign here”, “strike this clause” or “fill this field”, you do not need a text editor at all. Preview on macOS, the reader built into Microsoft Edge and the one in Firefox all add text boxes, highlights and freehand drawing, and they save back into the same PDF for free.

The catch is that annotations sit on top of the document rather than in it. That is fine for a signature and dangerous for anything you meant to remove: a white rectangle drawn over a paragraph hides it on screen only. The text underneath is still in the file and comes straight back out with copy-paste, a text extractor, or a search index. For anything sensitive, use a tool with an actual redaction feature — one that deletes the underlying content — or use the next method, where the passage you removed was simply never written.

Method 3: rebuild the PDF from its text

This is the route people rarely consider, and for a document that carries text rather than design it is usually the fastest and always the cleanest. Instead of fighting the page, you take the words out, edit them where editing is easy, and make a new PDF.

1. Open the PDF, select everything with Ctrl+A (Cmd+A on a Mac), and copy.
2. Paste into any plain-text editor — Notepad, TextEdit in plain-text mode, VS Code — and make your changes there. Find-and-replace works across the whole document, no font gets substituted, and nothing overlaps anything.
3. Paste the edited text into the text to PDF converter and download the result. The conversion runs in your browser, so the document you just edited is not uploaded anywhere.

What you gain: one consistent font, correct wrapping, page numbers if you want them, a file that is usually a fraction of the original size, and text that stays searchable and selectable. What you lose: the original layout, completely. That is the whole trade, and it is a good one precisely when the layout was never doing any work.

Worth rebuilding: meeting notes, transcripts, chat and email exports, log files going into a ticket, code listings, terms and conditions, documentation, book chapters, anything that started life as plain text and got printed to PDF along the way.

Not worth rebuilding: invoices, forms, certificates, anything with a letterhead, multiple columns, tables or images. Strip the layout out of those and you have destroyed what made them readable — edit those in place with Method 1.

If the text you want is already saved on disk rather than sitting on your clipboard, the text file to PDF converter takes up to 20 files at once and can merge them into a single document, each file starting on a fresh page.

Why this is harder than it should be

Inside a PDF, a page is a content stream: a sequence of operators that set a font, move to a coordinate and draw a run of glyphs. A run is not a word and not a sentence — it breaks wherever the typesetter adjusted spacing, so a single line of prose can be a dozen separate draws, each with its own offset. The file stores no information about where a paragraph starts or ends, because a printer never needed to know.

Every in-place editor therefore has to guess that structure back: group nearby draws into lines by their vertical positions, group lines into paragraphs by their spacing and indentation, then work out how to re-flow them when your edit changes the width. Add subset fonts with custom encodings, compressed object streams and a cross-reference table that has to be rewritten on save, and the surprise is not that PDF editing is fiddly — it is that it works as well as it does.