Can I crop individual pages differently?

Yes. You can crop specific pages with custom dimensions, or apply the same crop to all pages at once.

What units do the margin inputs support?

You can enter measurements in millimeters (mm) or as a percentage (%) of the page dimensions.

Does cropping reduce file size?

Yes. Cropping removes content and can significantly reduce file size, especially for scanned documents with large margins.

Will I lose text or images inside the crop area?

The cropped area is kept unchanged, cropping only adjusts the page boundaries, so any text or images inside the new boundaries remain selectable and intact. Only content outside the crop is removed.

Is there a file size limit?

PDFs up to roughly 100 MB work well on most devices. Because everything runs in your browser, the real ceiling is your device's available memory, larger files may be slower rather than failing.

Crop PDF Online

Remove unwanted margins, trim pages, or resize PDF content. Crop specific pages or apply to the entire document. Instant results in your browser.

Your files never leave your device

Drop PDF here or click to browse

Supports PDF files · up to 100 MB

How It Works

Upload a PDF file using the drop zone or file browser.
Preview the page and adjust the crop area by dragging edges or entering precise margin values.
Choose to apply the crop to all pages or just the current page.
Click 'Crop PDF' to apply the changes and download your cropped document.

When Do You Need PDF Cropping?

Cropping PDFs is essential for removing excessive margins from scanned documents, resizing pages to fit specific dimensions, or preparing documents for printing or digital sharing. Whether you have a document with large white borders or need to standardize page sizes across a batch of PDFs, this tool handles it instantly in your browser.

Cropping Options

Manual Margins · Enter precise margin values in millimeters to control exactly how much content to remove from each edge.
Visual Cropping · Drag the crop box edges directly on the preview to adjust the crop area interactively.
Preset Options · Use quick presets like "Remove Margins" or "A4 Center Crop" for common cropping scenarios.
Per-Page or Batch · Apply different crops to individual pages or the same crop to your entire document at once.

Frequently Asked Questions

Can I crop individual pages with different dimensions?

Yes. Select "Apply to current page only" and set your crop dimensions, then navigate to another page and apply a different crop if needed.

What is the maximum file size?

PDF files can be up to 100 MB. Processing happens in your browser, so very large files may take a few moments depending on your device.

Will cropping affect the quality of my PDF?

No. Cropping is a non-destructive operation that simply removes margins and adjusts page boundaries. All remaining content maintains its original quality.

Can I preview the crop before applying it?

Yes. The preview canvas shows your PDF page with the current crop area highlighted. Adjust the crop box or enter margin values to see the result in real-time before downloading.

Is my PDF uploaded to a server?

No. All cropping happens locally in your browser. Your PDF never leaves your device, ensuring complete privacy and security.

What "PDF crop" actually means

The phrase "crop a PDF" sounds like the operation crops an image: pixels outside the rectangle are discarded, the file shrinks, and the original content is gone. That is not what a PDF crop does. A PDF crop changes one of the five page-bounding rectangles in the page dictionary. The content itself, which is a stream of drawing operations (move pen, stroke, fill, place glyph, draw image), is preserved exactly as it was. The crop only changes which rectangle of that content the reader is supposed to display. Open the cropped file in a reader that ignores the rectangle and the original margins reappear unchanged.

This matters for three practical reasons. First, file size is essentially unchanged after a crop; the bytes representing the "cropped-away" content are still in the file. Second, the original margins can be recovered by any reader or tool that resets the relevant page box. Third, text that the eye-cropped portion contains is still selectable, copyable, and indexable by search engines, even though it does not render on screen. If the goal is to permanently delete margin content (for instance, a confidential header that must not be recoverable), the right operation is redaction followed by a re-rasterise, not a crop.

How this tool works

This tool combines two open-source libraries that have become the de-facto standards for in-browser PDF work. pdf-lib by Andrew Dillon handles structural operations: loading a PDF into an object graph, navigating page dictionaries, and writing a modified PDF back out. pdf.js by Mozilla handles the rendering side: drawing each page to a canvas so you can see the page and position the crop rectangle visually. Together they let the entire crop operation run inside the browser tab without any server contact.

When you press Crop PDF, the tool calls page.setCropBox(x, y, width, height) on each target page through pdf-lib. The arguments are in PDF user-space points, with the origin at the lower-left of the MediaBox. The tool converts your margin percentages to the page's actual MediaBox dimensions, then to the lower-left/upper-right convention pdf-lib expects. The modified PDF is serialised through pdf-lib's save(), wrapped in a Blob, and offered to the browser as a download. No bytes leave the tab during the operation. The roughly 1.3 MB pdf-lib bundle and the roughly 800 KB pdf.js bundle are cached after the first visit, so the tool loads instantly on return.

The five page boxes in the PDF specification

Every PDF page dictionary can carry up to five rectangles, defined in ISO 32000 section 14.11.2. Each rectangle is four numbers in default user-space units (one unit equals 1/72 inch), encoded as [llx lly urx ury] where ll is the lower-left corner and ur is the upper-right corner. Coordinates increase up and to the right, so the lower-left of a US Letter page is [0 0] and the upper-right is [612 792]. This tool sets the CropBox; the other four are left untouched.

MediaBox. The physical page boundary, mandatory on every page. This is the size of the sheet of paper the page would be printed on, or the canvas on which all other boxes are defined. The MediaBox is the only box that must exist; the other four inherit from it if absent.
CropBox. The visible region the reader displays and prints by default. If absent, defaults to the MediaBox. Cropping a PDF for screen-reading or general cleanup almost always means setting the CropBox, which is what this tool does.
BleedBox. The clipping region for production printing, including print bleed. Used by prepress workflows; ignored by most consumer PDF readers. Defaults to the CropBox.
TrimBox. The intended dimensions of the finished page after trimming. Production-printing context. Defaults to the CropBox. Note that "trim" in production printing is a different operation from the "trim margins" workflow that consumer tools call "crop."
ArtBox. The extent of the meaningful content on the page, excluding extraneous white space. Used as a hint by some layout applications. Defaults to the CropBox.

The PDF coordinate system, and why margin percentages are easier

PDF coordinates put the origin at the lower-left of the page and increase upward, which is the opposite of screen coordinates (origin at upper-left, y increases downward). The tool surfaces margin values from the top to match user expectations ("trim 5% from the top"), but internally flips the y-axis to translate that into the lower-left-anchored rectangle pdf-lib expects. Getting this wrong by a single flip is the most common bug in homemade PDF crop scripts: the cropped output ends up showing the bottom of the page when the user asked to keep the top, or vice versa.

The percentage-based margin inputs handle international page sizes uniformly. A4 paper is 210 x 297 mm, which is 595.276 x 841.89 PDF points. US Letter is 8.5 x 11 inches, which is 612 x 792 points. Some PDF generators produce non-standard sizes (Legal 8.5 x 14, A3, custom sizes from scanned documents). The tool reads the MediaBox dimensions from each page and computes the crop rectangle from the percentages, which works for any page size including the heterogeneous mixtures sometimes found in scanned documents.

Crop versus resize, trim, redact, and rotate

Several PDF operations sound similar but produce different results. Picking the wrong one is the source of most "the file did not change the way I expected" support tickets.

Crop. Sets the CropBox to a smaller rectangle. Content is preserved; file size is essentially unchanged; original margins are recoverable. This tool.
Resize. Changes the MediaBox dimensions, typically scaling the content to fit. The actual sheet size changes. Used when a document needs to render on a different paper size and the content should scale rather than be clipped.
Trim. In production printing, sets the TrimBox to the intended final cut size after bleed. Confusingly, many consumer tools use "trim" as a synonym for crop. In this tool's context, "trim margins" is the user-facing label for cropping away white margins.
Redact. Permanently removes content from the file: the bytes that drew the redacted region are stripped from the content stream and replaced with a black rectangle or removed entirely. The redacted content cannot be recovered. Required when hiding confidential information; a crop does not suffice.
Rotate. Changes the page rotation flag (the Rotate entry, 0/90/180/270). Does not alter the content stream; the reader applies the rotation at display time. Conceptually similar to cropping in that the underlying content is unchanged but the displayed result differs. Use PDF Rotate for that.

Real-world workflows that drive PDF cropping

Scanned documents with large white margins. The scanner placed the document on a flatbed sized for letter paper, but the document itself is half that size. Three quarters of each page is blank scanned-bed white. Cropping to the actual document boundary makes the file readable on smaller screens and reduces effective page area for printing.
E-reader and tablet reading. PDFs designed for letter paper at arm's length display tiny text on a 7-inch tablet because the reader fits the entire page width into the screen. Cropping out the margins recovers about 25% of the visible width and makes the text legible at a comfortable reading distance. Marvin, KOReader, ReadEra, Kobo, and the ReMarkable tablet all respect the CropBox; setting it once in this tool produces a permanent reading-optimised copy.
Print preparation. A document scanned at A4 needs to print on US Letter, or vice versa. Cropping to a centred A4 area inside a Letter MediaBox lets the document print correctly on Letter-stock printers without the auto-fit shrinking the page by ~3%. The "A4 Center Crop" preset performs exactly this conversion.
Form printing and form-only output. Single-page PDF forms designed for two-sided layout often have a top margin that contains form-prompt text and a bottom margin that contains submission instructions; the body of the form is in the centre. Cropping the form to the body region produces a form-only output suitable for digital fill-in or for printing a single side.
Slide-deck export cleanup. A presentation exported to PDF from PowerPoint or Keynote includes the slide framing on each page (notes area, footer, page numbers). Cropping to the slide content alone produces a clean PDF for slide-sharing or embedding in other documents.
Layout normalisation across multi-source documents. A merged PDF assembled from sources with different page sizes (an A4 cover from one source, US Letter body from another, a Legal appendix from a third) reads strangely because each section reflows the reader's zoom. Cropping every page to a uniform rectangle produces a visually consistent reading experience.
Hiding unwanted footers or headers before sharing. A document carries an organisation footer ("Internal use only, draft") that should be hidden when sharing externally. Cropping the footer region off produces a sharable version. Important caveat: the footer text is still in the underlying content stream and can be recovered by anyone who resets the CropBox. For permanent removal, redaction is required.

Common pitfalls and what they mean

Cropped PDF looks fine in Adobe Acrobat but shows the original margins in another reader. The other reader is ignoring the CropBox and falling back to the MediaBox. Browser PDF viewers in some configurations, command-line tools like pdftotext for layout-preserving extraction, and certain mobile apps fall into this category. The fix is to ensure the consuming reader respects the CropBox (modern readers do), or to run the output through a "flatten" pass in another tool to physically remove the cropped content.
The cropped PDF is almost the same size as the original. Yes. Setting the CropBox does not delete content. If the goal is a smaller file, the right tool is the PDF Compress tool, which downsamples images and re-compresses streams. Run this crop first, then compress.
Text that should be hidden is still selectable after the crop. Yes; the content stream is preserved. The text outside the visible rectangle is still indexable and copyable. Use redaction (a different operation) for permanent removal.
The crop rectangle drawn on the preview does not match the visible result. Most often a coordinate-flip bug in some other tool, but in this tool it indicates a non-standard page rotation. PDFs with a Rotate entry of 90 or 270 degrees have content drawn at the same coordinates but rendered rotated; a mismatch can arise if the PDF encodes rotation in the content stream rather than the Rotate entry. Reset rotation with the PDF Rotate tool and crop the normalised document.
A multi-source document with mixed page sizes crops wrong on some pages. "Apply to all pages" uses the same rectangle (in points) for every page. For documents with heterogeneous page sizes, switch to "Apply to current page only" and crop each section appropriately. Or normalise page sizes first with a resize-to-uniform tool.
Cropping a PDF with form fields appears to clip the fields. Form fields have absolute positions on the page. Cropping does not move them. A field crossing the CropBox boundary will render only the portion inside the box; the field is still functional and the data is intact, only the visible rendering is partial.
Annotations outside the crop region disappear from the page but stay in the comments pane. Annotations have absolute positions like form fields. Cropping hides them from the visible rendering but preserves them in the file. They reappear in any reader that ignores the CropBox and in the comments pane of any reader that lists annotations separately.

Browser-only versus cloud crop

The cloud PDF-crop services that fill the top of search results (Smallpdf, ILovePDF, PDF24 web, Sejda, Adobe Acrobat Online) all upload your PDF to their servers and run the crop server-side. The privacy implications are the same as any other cloud upload: the file transits the operator's network and exists briefly in operator logs and memory. The major operators publish privacy policies committing to deletion within hours and TLS in transit, and they have strong commercial reasons to comply, but "deletion within hours" is not "never seen."

This tool does not upload the file. The crop runs entirely in the browser tab using pdf-lib and pdf.js. You can prove it: open the browser developer tools to the Network tab before clicking Crop PDF, run the operation, and observe that no request fires with your file content. The trade-off is feature scope. Cloud services often offer "auto-detect margins" through OCR-style content-boundary detection, useful for batch work on hundreds of scanned pages. This tool offers a manual margin input plus three presets ("Remove Margins" for a small uniform inset, "A4 Center Crop" and "Letter Center Crop" for centred crops to standard paper sizes). For documents where the right crop is obvious by inspection, manual is faster than waiting for auto-detect; for documents where the content boundary varies across pages, auto-detect would save time.