Crop PDF pages
in your browser.

Trimming the edges of every page

The reasons to crop a PDF are usually about making the visible area match what should actually be looked at. A scan from a flatbed scanner picked up two centimetres of grey table around each page — that needs to come off before the document can be printed cleanly. A presentation exported from slide software has the speaker's notes column in a margin that should be hidden in the handout version. A scientific paper saved with conference watermarks in the bottom margin needs the bottom 30 mm shaved away to circulate as a clean preprint. A receipt scanned with a phone caught the edge of the desk; the kept part is a strip in the middle of each page. The job here is to pick what to keep and write a new page boundary around it.

The original page content stays — text, images, layout don't change. What changes is the rectangle the viewer treats as the page. Anything that fell outside the new rectangle is no longer drawn; the crop applies uniformly to every page.

What the four numbers mean

You give four margins — top, right, bottom, left — measured in PDF points from each edge inward. Points are the unit PDF uses internally: 72 points per inch, which works out to roughly 2.835 points per millimetre. An A4 page is 595 × 842 points. Setting top to 36 trims half an inch off the top of each page; setting all four to 36 trims a half-inch frame from the whole document.

Three rules:

• Margins must be non-negative. You can't grow the page with a negative number — that's outside the original content. Use a viewer to add whitespace around a PDF instead.
• The trimmed area can't exceed the page. If left + right would equal or exceed the page width, the tool stops with "margins exceed page dimensions". Same for top + bottom and the height.
• The same numbers apply to every page. If your document has a mix of A4 and Letter pages, a 36-point crop trims both consistently — but if you need different crops on different pages, that's not in this flow yet.

Three presets for common cases

For the most common shapes, presets do the math:

• Remove margins. Trims 36 points (≈12.7 mm, half an inch) from every side. Standard for "make this scan fit the page without grey borders" cases. Good first try if you're unsure.
• Square. Crops each page to a centred square using the shorter dimension. A4 (595×842) becomes 595×595, kept centred horizontally with a 123-point trim above and below. Useful for portfolios where slides need to fit Instagram-style frames or document scans need consistent aspect ratios.
• A5 from A4. Keeps the left half of each page, dropping the right. The classic "I have a two-up A4 export and want the left page" case — your A5 reading copy comes out without re-exporting from the source.

If you set a preset, the four manual margin numbers are ignored. To combine "remove margins" with custom additional trim, run the tool twice — once with the preset, once with the extras.

How the page boundary is changed

PDF has two boundary boxes that matter here. The MediaBox is the physical sheet — what a printer would see. The CropBox is the visible region — what a viewer shows when reading. This tool updates both. That matters because a CropBox-only crop is just a hint: another PDF tool can ignore it and reveal the original sheet again. Updating the MediaBox makes the crop permanent — viewers, printers, and other tools all treat the new rectangle as the entire page.

The content outside the new boundary is still in the file, technically — pdf-lib doesn't re-encode the page content stream. But every viewer treats it as cropped, and re-saving in any other tool will discard the off-page content for good. For most purposes this is "the crop is permanent"; for forensic purposes ("is this content really gone?"), it isn't, and you'd want a tool that re-rasterises pages.

What survives and what doesn't

• Text, images, fonts, layout — all preserved. The page content stream isn't touched. Selectable text remains selectable; embedded fonts stay embedded.
• Bookmarks, annotations, links, form fields all survive. Annotations whose bounding rect was inside the kept area still show; ones in the trimmed margin are hidden by the new boundary but still exist in the file.
• Page numbers, headers, and footers are part of the page content. If your trim catches the footer, the page number that was there is gone visually — the page-numbering tool re-adds them inside the new boundary if you need.
• Digital signatures are invalidated. Any modification to a page breaks the signature binding.

If something looks off

• "Crop margins exceed page dimensions." The numbers add up to more than the page is wide or tall. Halve them, or work in millimetres mentally — 36 points ≈ 12.7 mm, so for a 595-point-wide A4 you have about 210 mm of room.
• The crop took content I wanted to keep. Re-run with smaller margins. The original file isn't modified — what you downloaded is a separate file with the trim baked in.
• Pages with different sizes cropped inconsistently. Expected. The same numbers apply to every page; an A4 + Letter mix gets the same edge-inward trim, which leaves different visible dimensions. Split the document first via split-pdf, crop each part, then merge with merge-pdf.
• The PDF is encrypted. Run it through unlock-pdf first; encrypted PDFs can't be modified.

Practical notes

1. Measure first when you're not sure. Open the PDF in a viewer that shows page dimensions — the numbers are usually in the export dialog. Note where the unwanted content sits, then convert to points (mm × 2.835) before entering values.
2. For trimmed scans, "remove margins" preset handles 90% of the cases. If you still see grey edges after, re-run with custom values around 50–60 points instead of 36.
3. Two-up exports → single pages. "A5 from A4" preset gives you the left half; to get the right half, set left = width / 2 manually. For a 595-point-wide A4, that's left = 297.
4. The original file stays untouched. What you download is a new PDF with the smaller pages; the source on your disk is unchanged.

What happens to your file

Cropping runs in your browser. Open DevTools and watch the Network tab during the operation — there are no outbound requests carrying the file content. The PDF stays on your disk; the cropped version is a new download alongside it.