How to Check PDF Page Size in Points and Catch Mixed Pages Before Printing

A PDF that looks fine on screen can still be the wrong size for the printer. The page boundary lives in the file as a small set of numbers, and those numbers are what a print shop, an e-sign portal, or an upload form actually reads. When one page in a fifty-page packet carries a different boundary than the rest, nothing on screen warns you. The job just comes back scaled, cropped, or rejected.

This guide walks through how PDF measures a page, how to read those measurements, and how to find the one odd page hiding in a long document before it costs you a reprint.

PDF measures pages in points, not pixels

A PDF does not store its page size in millimeters or inches. It stores it in points, and the PDF point is fixed: 72 points equal one inch. That number never changes with zoom, screen resolution, or the device you open the file on. It is baked into the file format.

Once you know the rule is 72 points per inch, the common sizes turn into round, recognizable numbers:

A4 is 210 mm by 297 mm, which works out to 595 by 842 points.
US Letter is 8.5 by 11 inches, which is 612 by 792 points.

So when a page reports a boundary of 595 by 842, that is A4. A boundary of 612 by 792 is Letter. The two sizes are close enough that a person flipping through the document will never notice the difference, but the raw numbers make it obvious in a second. The width gap is only 17 points, about a quarter inch, and the height gap is 50 points, but a press treats them as two entirely different stocks.

Two boundaries are worth knowing about. The MediaBox is the physical page edge. The CropBox is the visible region many viewers and print workflows use for display. A file can have a tidy MediaBox and a CropBox that quietly trims part of the page, so it pays to look at both rather than trusting what shows up on screen.

Why a 50-point mismatch breaks a print job

Checking the sizes catches mixed or oversized pages that would otherwise print wrong. A press, a binder, or an imposition step expects one consistent stock. Feed it a packet where page 14 is Letter and everything else is A4, and one of two things happens. Either the press refuses the file outright, or it scales the odd page to fit, which shifts every margin, nudges text toward the trim edge, and leaves you with a page that does not match its neighbors.

The same logic applies to oversized pages. A scan saved at an unusual capture size, a slide exported at 1280 by 720 points, or a poster page accidentally merged into a report will all sail through a casual on-screen review and then blow up at the print stage. Reading the dimensions up front turns a guess into a fact.

A worked example: one Letter page in an A4 contract

Here is the situation that sent me looking for a faster way to do this. I had a 32-page contract assembled from three separate exports, and the print vendor bounced it with a note that said "inconsistent page size." Opening the file and scrolling through it told me nothing. Every page looked identical at a glance.

So I ran the PDF through the PDF Page Size Inspector instead. It reads the MediaBox and CropBox declarations, groups repeated sizes together, converts the points into inches and millimeters, and shows how many times each size appears. The result was three rows:

595 x 842 points (A4), portrait, 31 pages
612 x 792 points (US Letter), portrait, 1 page

There it was. One page, somewhere in the middle, had come from a colleague's Letter-default template. The count made the search trivial: I did not have to inspect 32 pages, I had to find the single Letter page and re-export it at A4. The vendor accepted the corrected file the same afternoon.

That count column is the part that saves the most time. A first-page-only check, which is what most people do by reflex, would have reported "A4" and sent the broken file straight back to the press.

How to read the size report

When you inspect a PDF this way, you get a small table instead of a wall of pages. Read it like this:

Look at the row count first. One row means every page shares a boundary. Two or more rows means the document is mixed, and the page counts tell you how lopsided the split is.
Match the numbers to a known size. 595 by 842 is A4, 612 by 792 is Letter, 792 by 1224 is Tabloid, and so on. Anything that does not land on a familiar size is worth a second look.
Compare MediaBox to CropBox. If they differ, part of the physical page is being hidden from view. That is fine on purpose and a problem by accident.
Check orientation. A landscape page mixed into a portrait run is another classic culprit, and the report labels each row portrait or landscape so it stands out.

If you also need to confirm a file is genuinely a PDF before you trust any of these numbers, the File MIME Type Inspector reads the magic bytes so a mislabeled or renamed file does not waste your time.

The inspection runs locally

Contracts, scanned IDs, sealed bids, and unreleased designs are exactly the kind of documents you do not want to upload to a stranger's server just to read their page dimensions. There is no need to. Page size lives in structural box declarations near the top of each page object, so the inspector reads that data directly in your browser. The file never leaves your machine, nothing is rendered, and nothing is uploaded. You get the numbers and the document stays where it belongs.

A quick checklist before you send a PDF to print

Confirm every page shares one MediaBox, not just the first page.
Translate the points to a named size: 595x842 for A4, 612x792 for Letter.
Watch for a CropBox that differs from the MediaBox.
Flag any landscape page sitting inside a portrait document.
Re-export, do not rescale, any page that does not match.

Five seconds of reading the size table is a far better trade than a reprint or a missed submission deadline. The numbers are already in the file. They just need someone to look at them before the press does.

Made by Toolora · Updated 2026-06-13