Why is my PDF so big?

It is almost always the images

Text is astonishingly cheap to store. A page of characters, with the font already embedded, costs a few kilobytes. Anything that surprises you about a PDF's size is therefore almost never the words.

Before changing anything, it is worth knowing which of the five usual causes you have — because the fix for one does nothing for another, and the wrong fix costs quality without recovering much.

The five causes

Ranked roughly by how often they turn out to be the answer:

• It is a scan. Every page is a photograph. A 20-page colour scan at 600 DPI can pass 100MB, and nothing about that is unusual — it is what the file is. Re-encoding the images at a sensible resolution is the fix, and it is dramatic.
• Images embedded at full camera resolution. A phone photo dropped into a document is stored at its original 4000-pixel width even though it displays two inches wide. The page shows a thumbnail; the file carries the original.
• Print-resolution export. Design tools default to 300 DPI and no image downsampling because the output might go to a press. For a file that will only ever be read on screen, that is several times more data than anyone will see.
• Embedded fonts, not subsetted. A font file embedded whole rather than trimmed to the characters actually used can add a megabyte or two per font. Uncommon in modern exports, but it happens with older tools and with CJK fonts, which are enormous.
• Revision history. Some editors save changes by appending rather than rewriting, so a file edited twenty times can contain twenty versions of itself. The page count stays the same and the size quietly triples.

Telling a scan from a text document

This single test decides most of what follows. Open the file and try to select a line of text with your cursor.

If the text highlights, you have a text PDF. The size is coming from embedded images, fonts or revision history, and compression will help in proportion to how many pictures there are. A text-only report that will not shrink is simply already efficient — that is the honest answer, and squeezing it further only degrades it.

If your cursor draws a box over what is obviously a picture, it is a scan. The images are the entire file, compression works extremely well, and you should expect to remove most of the size without a visible difference.

What compression actually does

Worth understanding, because "compress PDF" describes two quite different operations and the difference decides whether your text survives.

The good version re-encodes the embedded images: it downsamples anything stored at more resolution than the page uses and re-compresses it at a lower quality, leaving text, vectors and structure untouched. Your document stays selectable and searchable and just gets smaller.

The crude version renders every page to a bitmap and rebuilds the document from those pictures. It compresses harder, and it turns your text into a photograph of text — no longer selectable, searchable or accessible to a screen reader. For a scan that is harmless, because the pages were already images. For a contract it is a significant loss, and tools that do it rarely say so.

Preventing it at the source

If you are the one producing the file, most of this is decided at export. Choose a screen or web preset rather than press quality unless it is genuinely going to a printer. Resize large photographs before placing them rather than after. And if a document has been edited many times, exporting a fresh copy rather than saving again discards the accumulated history.

Frequently asked questions