How to add a watermark to a PDF (CONFIDENTIAL, DRAFT, or your own text)

Adding "CONFIDENTIAL" or "DRAFT" across every page of a PDF is one of those workflows that sounds simple and isn't, depending on which tool you have. Adobe Acrobat Pro has a dedicated Watermark feature. Free PDF readers don't. Online watermark services ask you to upload your document. The native macOS Preview app doesn't have a watermark function. For internal documents that need a handling indicator before they're shared, the friction is enough that most people skip it.

The cases where you actually want a PDF watermark

Confidentiality marking. A "CONFIDENTIAL" or "INTERNAL USE ONLY" mark on every page tells recipients (and accidentally-forwarded copies) how the document should be handled. Survives screenshots and partial-page copy-paste.
Draft / non-final indication. A "DRAFT" watermark prevents people from treating a working document as the final version. Critical for legal drafts, contract negotiations, anything where an unfinished version going out as final is a real risk.
Attribution. A subtle footer with the source name or copyright. Less intrusive than full-page diagonal watermarks.
Tracking-style watermarks. A unique recipient name or ID on each copy so leaked versions can be traced. This requires a per-recipient batch workflow; the tool here covers the same-watermark-everywhere case.

Step-by-step with our tool

Use PDF Watermark to add a text watermark to every page of a PDF, entirely in your browser. The tool uses pdf-lib to draw text into the page's content stream, so the watermark becomes part of the page itself — it shows up in every viewer, on print, on screenshot, everywhere a page is rendered.

Open justfiletools.com/tools/pdf-watermark.
Drop a PDF.
Enter your watermark text. CONFIDENTIAL, DRAFT, © 2026 Your Company, or whatever else fits.
Pick a position. Four options cover the common cases:
- Diagonal across the page — the stock-photo-watermark look, rotated -45° by default. Hard to crop out. Right pick for "CONFIDENTIAL" on documents that may get screenshotted.
- Center horizontally — non-rotated centered text. Good for short marks like "DRAFT".
- Top of page — header-style mark above the content.
- Bottom of page — footer-style mark. Least intrusive.
Adjust the appearance: color (defaults to gray, but bright colors work for warnings), font size (12 to 200 px, default 60 for diagonal mode), opacity (default 0.2 — subtle but visible). Rotation slider for diagonal mode.
Click Apply watermark. The output downloads as a new PDF with the watermark baked into every page.

How the rendering works

pdf-lib's drawText method adds text operations to the page's content stream. The text is rendered with Helvetica Bold from the PDF standard 14 fonts (every PDF viewer has these built in, so no font embedding is needed and the output file size is barely larger than the input). The watermark text becomes part of the page just like the original content; PDF viewers render it in the same compositing pass, no separate "watermark layer" that can be toggled off.

The diagonal position uses pdf-lib's rotation primitive to rotate the text around its placement point. Rotation is in degrees; -45° is the classic stock-photo angle. Opacity is applied via the PDF spec's opacity property — every modern PDF viewer respects it, producing the see-through effect.

Things to watch out for

Watermarks can be removed. Watermarks are part of the page's drawing operations. A determined recipient with a PDF editor can remove individual drawing operations — including watermarks — in minutes. So watermarks are friction, not security. For real anti-tampering, you need cryptographic signing (PDF digital signatures) which proves provenance even if a recipient re-saves a modified copy.

The watermark adds to file size, slightly. Each page gets a small text drawing operation added to its content stream — roughly 100–200 bytes per page. A 100-page PDF gets a ~20 KB increase. For practical purposes, the watermarked file is the same size as the input.

Pages of different sizes are handled correctly. If your PDF mixes Letter and A4 pages, each watermark is placed correctly for that specific page's dimensions. The position calculations use each page's own width and height.

Encrypted PDFs are not supported. Decrypt the input first; pdf-lib can't apply watermarks through encryption.

Choosing position and opacity

The defaults work for most cases. Diagonal at 0.2 opacity is subtle but visible — anyone who looks for the watermark sees it, but it doesn't visually dominate the document content. For more intrusive marking (where you want the watermark to clearly signal "do not redistribute"), bump opacity to 0.4–0.6. For very subtle attribution (where you want the watermark only when someone looks closely), drop to 0.1.

For document content that has lots of dark areas (charts, dark images, photo-heavy documents), the default gray watermark may be hard to see. Switch to a bright color (red, yellow) and increase opacity. For documents with mostly light backgrounds (text-only contracts, white-background reports), the default gray at 0.2 is exactly right.

Alternative approaches and when to use them

Adobe Acrobat Pro. The most full-featured watermark tool — supports image watermarks, watermark with PDF templates, per-page customization, and removal of existing watermarks. $20/month.
LibreOffice Draw. Free, opens PDFs, you can add text or images over pages and re-export. Works but is more friction than a dedicated tool.
pdftk + LaTeX. For batch workflows where you want each PDF in a directory to get the same watermark, the pdftk input.pdf stamp watermark.pdf output watermarked.pdf workflow is scriptable. Requires preparing a watermark template PDF first.
Online watermark services. Fast but upload-based. Wrong tradeoff for sensitive documents (a watermarked-as-CONFIDENTIAL document is, by definition, not something you want sitting on a stranger's server).

Privacy considerations

The whole point of a CONFIDENTIAL watermark is that the document is sensitive — so uploading it to a watermarking service to add the watermark is a contradiction. The in-browser tool resolves this by never sending the PDF anywhere. pdf-lib loads the PDF into your browser's memory, applies the watermark in JavaScript, and outputs a blob URL. Network panel stays empty. The watermark is added by the same browser tab that loaded the PDF; nothing else sees it.

Related tools and guides

PDF Watermark — the tool this guide covers.
Image Watermark — the same operation for JPG/PNG images.
PDF Metadata Viewer — strip identifying metadata before sharing.
PDF Redaction — for hiding (not just marking) sensitive regions.
How to redact a PDF safely.