What "cropping" a PDF actually means
Cropping a PDF is not what most people think it is. When you crop a JPEG, the pixels outside the crop are gone — the file gets smaller and the discarded area is unrecoverable. PDFs work differently. Every page has two rectangles attached to it: the MediaBox (the page's full physical size, like the sheet of paper) and the CropBox (the visible region readers display, like a window onto the sheet). Cropping a PDF means shrinking the CropBox. The MediaBox stays the same. The content outside the new CropBox is still in the file. It is just not shown.
This is non-destructive cropping, and it is the right default for most PDF work. You can un-crop later. You can extract the hidden margins if you discover you cropped too aggressively. The downside: file size does not shrink meaningfully, because the content is still there. The Crop PDF tool does non-destructive cropping. If you need to actually delete the cropped-out pixels (destructive cropping), you have to rasterize the PDF — and rasterizing loses text selection, hyperlinks, and form fields, which is rarely what you want.
How to use the Crop PDF tool
The widget asks for four numbers: how much to trim off the top, bottom, left, and right edges of every page. The numbers are percentages because PDFs come in every size imaginable — US Letter, A4, A6, custom scanner output, screen captures — and a percentage gives sensible results across all of them. A 10% top margin trims a meaningful slice off any page size.
- Drop or pick your PDF. Up to 100 MB. The tool reads the first page's dimensions so you can sanity-check the crop result before applying it.
- Use a preset (Tighten 5%, Tighten 10% all sides) or type per-edge percentages by hand. Zero means no crop on that edge; 50 is the maximum on any single edge.
- The widget shows the resulting page size as you type, in points and in inches/centimeters. Same crop applies to every page in the PDF.
- Click Crop PDF. Download. The output is named after your input plus "-cropped" (so
scan.pdfbecomesscan-cropped.pdf).
The file never leaves your browser. Crop PDF runs entirely on your machine via pdf-lib. Check your network tab during the crop operation — you will see zero outbound requests.
The honest version: non-destructive vs destructive crop
Most "Crop PDF" tools you find online (iLovePDF, SmallPDF, PDF24, Sejda) also do non-destructive cropping. They do not advertise that fact because "your file is smaller now" is the headline most users expect. The fact is, after a non-destructive crop, the file is the same size give or take a few bytes — only the CropBox rectangle changed. We tell you up front so you do not spend an hour wondering why your PDF did not shrink.
The simple rule: use Crop PDF when you want pages to display smaller (less wasted paper when printing, less wasted screen space when reading). Use Compress PDF when you want the file size to drop. The two operations sound similar and solve completely different problems.
If you really need destructive cropping — pixels gone, smaller file — the workflow is to crop with this tool, then rasterize the result using Compress PDF at a high enough setting that it bakes the visible page into images. You lose text selection in the output. There is no free lunch.
Worked example: trimming a scan
You have a 20-page contract scanned at 300 DPI. The scanner left roughly an inch of white margin on every side. Each page is US Letter (8.5 × 11 inches, or 612 × 792 points in PDF units). You want to trim the margins so the printed copy uses less paper and the on-screen view zooms in tighter on the actual text.
An inch on an 8.5-inch-wide page is about 12% of the width. On the 11-inch height, an inch is about 9%. So your crop numbers are 9% top, 9% bottom, 12% left, 12% right. Plug those in and the widget will show a new page size of roughly 6.6 × 9.0 inches — close to A5, much closer to the actual content.
| Edge | Original (US Letter) | After 9%/12% crop | What you see |
|---|---|---|---|
| Width | 8.5 in (612 pt) | 6.6 in (478 pt) | Margins gone; text fills more of the page |
| Height | 11.0 in (792 pt) | 9.0 in (648 pt) | Top/bottom whitespace gone |
| File size | ~3.5 MB | ~3.5 MB | Unchanged — pixels still there, just hidden |
Print this on US Letter paper and you will see the contract content scaled up to fill more of the sheet — useful for older readers or when you want to fit two pages side-by-side on one sheet. View it on a phone and the text is bigger because the reader no longer has to fit the original margins on screen.
Why percentages, not inches
The widget takes percentages for each edge instead of absolute measurements (inches or centimeters or points). The reason: a single number works across every page size, and the result is what you actually want.
If you set a 1-inch crop on every edge, the result is fine on US Letter — you lose about 12% of the width and 9% of the height, which sounds reasonable. Try the same 1-inch crop on an A6 page (4.1 × 5.8 inches) and you have cropped 49% of the width away. The page becomes a thin vertical strip. Try the same 1-inch crop on a tabloid page (11 × 17 inches) and you have only cropped 9% of the width — barely noticeable.
Percentages give you the same visual result on any page size. A 10% margin trim looks like a 10% margin trim whether the input is A4, US Letter, or a custom 8000 × 6000 scan. If there is enough demand for absolute units, we will add a unit picker. For most cropping work, percentages are the right answer.
Why this beats the upload-and-watermark crowd
The big online PDF tools do roughly the same thing under the hood (most use pdf-lib or a similar library; cropping is three lines of code). What they wrap around it is the problem: upload to a server you do not own, a 5 MB cap on the free tier, a watermark on the output unless you pay, a daily quota that resets in 24 hours and locks you out at the worst moment.
Cropping a PDF is not a remote-server operation. Your file does not need to round-trip through someone else's data center in another country. The Crop PDF tool uses the same pdf-lib library those services run on their servers, but it runs in your browser. Files never upload. No watermark. No daily limit. No signup. The only cost is that your laptop's RAM is doing the work instead of theirs, which is why the upper bound is 100 MB and not 100 GB.
What gets preserved
Because the crop is a CropBox metadata change and not a re-render, everything else passes through unchanged:
- Text stays selectable and searchable. Ctrl-F still works on the cropped output.
- Bookmarks and hyperlinks remain clickable.
- Form fields remain fillable. If your input had filled form values, they survive cropping.
- Annotations and comments pass through. Even comments anchored to coordinates outside the new CropBox stay attached to the document.
- Signatures stay valid. Both drawn-image and digital signatures pass through because the underlying content is not modified.
The contrast with destructive operations is sharp. Flatten PDF Form deliberately removes form-field interactivity (that is the point). Compress PDF at the aggressive end re-encodes images and can lose quality. Crop PDF touches one rectangle per page and leaves everything else alone.
Edge cases
- Some print drivers ignore the CropBox. Most modern drivers (CUPS on macOS/Linux, Windows native print, Adobe Reader's print path) respect the CropBox. A small number of older industrial print drivers use the MediaBox, which means they will print the uncropped page. If you crop, then print, then see the full original page on paper, this is the cause. The fix is to use a destructive crop workflow or to use a different print path.
- You want different crops on different pages. The current tool applies the same crop to every page. Per-page cropping with thumbnail previews is on the roadmap. Workaround today: Split PDF to separate the pages that need different crops, crop them independently, then PDF Merger to reassemble.
- You crop too aggressively and want to undo. Re-run the tool on the original input file with new (smaller) percentages. The tool does not modify the source file, so the original is always there. If you want to recover content from a cropped output where you no longer have the original, the data is technically still in the file (CropBox is non-destructive) but most readers will not let you see beyond it without manual editing in a desktop PDF editor.
- Password-protected PDFs. pdf-lib refuses to open encrypted PDFs. Remove the password first using a desktop reader.
- Crop values that sum past 100. Top + bottom must sum to less than 100, same for left + right. Setting top + bottom to 90% would leave 10% of the page visible — technically possible but rarely what you want. The widget caps each individual edge at 50% and warns if a pair sums to an unreasonable total.
Related tools
Cropping is usually one step in a longer PDF workflow:
- Rotate PDF — fix orientation before cropping, since margins are different on a sideways page.
- Split PDF — separate pages that need different crops.
- PDF Merger — reassemble after splitting and cropping individually.
- Compress PDF — actually shrink the file size after cropping for display.
- PDF Extract Text — pull just the text out if you need the content but not the page layout.
Frequently asked questions
What does "non-destructive" actually mean?
PDF pages have two main boxes: the MediaBox (the page's full physical size) and the CropBox (the visible region readers display). This tool updates the CropBox. The MediaBox stays the same — the content beyond the crop is still in the file, just not shown. Most readers (Acrobat, browsers, Preview) respect the CropBox. Some print drivers use the MediaBox and will print the uncropped version. For destructive cropping (actually removing pixels), you would rasterize the PDF first using Compress PDF, but you lose text selection.
Can I apply different crops to different pages?
Not in version one — the same crop applies to every page. Per-page cropping with thumbnail previews is on the roadmap. If you only need to crop one page, use Split PDF to extract it, crop it here, then PDF Merger to put it back.
Will the file size shrink after cropping?
Slightly — only by a few bytes (the size of the new CropBox descriptor). Because cropping only updates metadata, the underlying page content is unchanged. To actually shrink an image-heavy PDF, use Compress PDF instead. It is lossy but produces real size reductions on the right input.
Why percentages instead of inches or centimeters?
Percentages work across any page size. A 10% top margin gives a sensible result whether the page is US Letter, A4, or a custom 8000 × 6000 scan. With fixed units, a 1-inch margin trims a quarter of an A6 page but only a small slice of a tabloid. We will add a unit picker if there is demand for absolute measurements.
Is my PDF really not uploaded?
Correct. The crop runs in your browser via pdf-lib. Zero outbound requests during the operation. Check your browser's network tab to confirm. The Microapp page itself was served from our CDN; everything after page load happens on your machine.
Does it preserve bookmarks, links, and forms?
Yes. Crop only updates the CropBox — bookmarks, hyperlinks, annotations, form fields, and signatures pass through unchanged. The form values stay filled in. Hyperlinks remain clickable.
What about password-protected PDFs?
pdf-lib refuses to open encrypted PDFs. Remove the password first using a desktop PDF reader's "save as unprotected copy" feature, then run Crop PDF on the resulting file.
What is the maximum crop on any one edge?
Each per-edge crop is capped at 50%. Setting top + bottom to a combined 90% leaves only 10% of the page visible — possible but rarely useful. Top + bottom must sum to less than 100, same for left + right. The widget will not let you submit values that would crop the entire page away.