Have a PDF that opens sideways or upside down? ButDoc's Rotate PDF tool permanently fixes page orientation, so the document reads correctly every time it is opened, emailed or printed — not just while you are looking at it.
Select the pages that need turning, rotate left or right, and download. Below: why rotating in your PDF viewer never sticks, why this operation costs your document nothing in quality, and the scanner quirk that puts every second page upside down.
Almost everyone tries this first, and it is worth understanding why it fails.
The rotate buttons in Adobe Reader, Chrome's built-in viewer, Preview on a Mac and most other readers change how the page is displayed to you right now. They are a view setting, like zoom. Nothing is written back into the file. Close the document and the setting is gone; email it to a colleague and they see the original sideways page, because the file itself never changed.
What you need is rotation saved into the document's own structure. That is what this tool does. Once applied, the correct orientation travels with the file — to every reader, every device, every printer, and everyone you send it to.
This surprises people who expect rotation to degrade a document the way re-saving a JPEG does. It does not, and the reason is worth knowing.
Inside a PDF, every page carries a small piece of metadata — a rotation value that can only be 0, 90, 180 or 270 degrees. Rotating a page does not redraw it, re-encode its images, or reprocess its text. It changes that one number, and every reader that opens the file afterwards obeys it.
The practical consequences are all good:
The same constraint explains a limitation: because the value can only be a multiple of 90, rotation cannot straighten a scan that went through the feeder slightly crooked. That is a different problem — see below.
If you have ever scanned a double-sided stack and found the odd pages perfect and the even ones rotated 180 degrees, this is why.
Many document feeders scan one side, then flip the sheet along its short edge to capture the other. Physically, that turns the reverse side upside down relative to the front. The scanner captures exactly what it sees, so the file alternates: page 1 correct, page 2 inverted, page 3 correct, page 4 inverted.
The fix is straightforward once you know the pattern: select only the even-numbered pages and rotate them 180 degrees. Do not rotate the whole document, or you will simply move the problem to the odd pages.
You can also prevent it at source. Most scanner software has a duplex setting for flipping along the long edge rather than the short edge, or an option labelled something like "rotate back side 180°". Setting that once saves you the correction on every future scan.
If your PDF is a scan and you intend to make its text searchable, order matters.
OCR works by recognising letter shapes, and it expects those shapes the right way up. Most engines will attempt to detect and correct orientation automatically, but accuracy drops noticeably on sideways pages, particularly with small print, tables, or any language that does not use the Latin alphabet. Text can come out garbled, or a page can be skipped entirely.
Rotating first removes the guesswork. Fix the orientation, then run OCR PDF on a document where every page is already upright. The recognition quality is better and the result is more consistent.
The same applies to anything else that reads a PDF's content: text extraction, converting to Word, or copying and pasting from the document.
Three related problems look like orientation issues but need different fixes.
The page is slightly crooked, not sideways. A sheet that entered the scanner at a three-degree angle needs deskewing, not rotation. PDF rotation only works in 90-degree steps, so it cannot help. Rescanning is usually the fastest fix; some scanner software has automatic deskew.
The page prints sideways even though it looks correct. This is normally the print dialog rather than the document. Check for a setting such as "Auto-rotate and centre" or a page orientation override, and make sure the paper size matches the document's page size. A landscape page printed onto portrait paper will be rotated by the driver to fit.
The document mixes portrait and landscape deliberately. Reports often include wide tables or charts on landscape pages among portrait text. Those pages are meant to be that way. Rotate only what is genuinely wrong, which is why selecting individual pages matters more than a bulk rotate.
Working on a scan? OCR PDF makes scanned text searchable, Crop PDF trims uneven margins, Delete Pages removes blank sheets the feeder picked up, and Compress PDF shrinks a large scan for email. Browse all PDF tools.