Why Is My PDF So Large? A Practical Guide to PDF File Size
July 20, 2026 · PDF Tools
Two ten-page PDFs can differ in size by a factor of a hundred: one is 80 KB, the other 40 MB, and at a glance they look the same. The difference is never magic and almost never the text. Understanding what actually occupies the bytes in a PDF tells you immediately whether a file can be made smaller — and by how much.
What takes up space in a PDF
A PDF is a container holding several kinds of content, and they have wildly different costs. Text is nearly free: a page of prose is a few kilobytes, because the file stores character codes and positions, not pictures of letters. Fonts cost more — a document that embeds several full font families can carry a megabyte or two of font data — but the real weight in almost every oversized PDF is images.
Images dominate because of arithmetic. A single A4 page scanned at 300 DPI in colour is roughly 25 million pixels; even after JPEG compression that can be a megabyte or more per page. A 40-page scan at that resolution is a 40–60 MB file before anything else is counted. Office documents exported to PDF pick up weight the same way: every photograph, logo and chart screenshot is embedded at whatever resolution it had, which is often far more than the page will ever display.
The three common causes of a huge PDF
- Scans at needlessly high resolution: scanner defaults of 300–600 DPI in full colour are meant for archival reproduction, not for emailing a signed form. 150–200 DPI in greyscale is perfectly readable and a fraction of the size.
- Photographs inserted at camera resolution: a 12-megapixel phone photo dropped into a report is carried at full size even when it is displayed three centimetres wide. A handful of such photos adds tens of megabytes.
- Duplicated or unoptimised resources: the same logo embedded once per page, full font sets where a subset would do, or editing history some producers append to the file. Individually small, collectively significant.
What compression can and cannot do
PDF compression tools work almost entirely on images: they scale down pictures whose resolution exceeds what the page needs, and re-encode them with stronger JPEG compression. On a scan or a photo-heavy report this routinely removes 50–90% of the file. On a text-only contract it does nearly nothing — the text was already compact, and there is no fat to trim. This is why the honest answer to "how much smaller will it get?" is always: it depends on what is inside.
Compression is also lossy where it acts. Aggressive settings soften fine detail — small print, hairline table borders, the texture in a map. The right level depends on destination: a document read on screen tolerates far more compression than one going to print. Good practice is to compress a copy, open it, and check the smallest text before sending.
When compression is the wrong tool
If a file is large because it is long, splitting beats compressing: a 300-page board pack that exceeds a mail limit is better delivered as the twelve pages the recipient needs. If a file is large because it was scanned at heroic resolution, the best fix is upstream — rescan at sensible settings — because information discarded by a scanner's JPEG pass cannot be recovered later. And if a file has already been compressed once, compressing it again mostly degrades quality: redundancy, once removed, does not grow back.
A practical checklist
- Look at what the document is: mostly text, mostly scans, or text with photos. That predicts the outcome before you start.
- Compress a copy at a middle quality level and compare sizes.
- Check the output — smallest text, finest lines — before replacing anything.
- Still too big? Consider whether the recipient needs every page, and split instead.
- For scans, fix the source: 150–200 DPI greyscale is the sweet spot for ordinary documents.