How to Read a PDF on a Kindle or Kobo, and When to Convert It

The short answer

Match the path to the PDF. A novel or a report in one column can be converted with PDF to ePub. An article with two columns or a table is better left as a PDF. If its margins are wide, trim them first with Crop PDF. A scan needs OCR in its own language before anything else.

In our check on October 30, 2026, we ran three small test PDFs: a novel, a two-column article with a table, and a scan in French. We did not open any result on a Kindle or a Kobo, because we have no e-reader. What the files hold, we checked. What the readers do comes only from Amazon's and Kobo's help pages, quoted with the date we read them.

PDF to ePub, OCR PDF and ePub to MOBI run on our server. The uploaded file is deleted as soon as the job finishes, and the result is wiped within about two hours. Crop PDF runs in your browser. All four are free and need no account.

What Amazon and Kobo say they accept

We read the pages below on October 30, 2026. Amazon's help page on sending documents to your Kindle library lists what Send to Kindle takes. The list includes “PDF (.PDF)” and “EPUB (.EPUB)”. For the web upload it says: “Upload documents (200 MB or smaller) directly from your device with Send to Kindle for Web.”

Kobo's help page on file formats lists PDF and ePub for its eReaders. It adds a note on ePub 3: “Kobo Original and Kobo Wi Fi does not support EPUB3”. Our converter writes ePub 3, so check your model. The list for the Kobo Books app on Android has no PDF.

Another Kobo page, “Turn pages, magnify, and rotate PDFs using your Kobo eReader”, lists what you can't do with a PDF book. It includes “Change the text size or style”. For a novel, that is our reason to convert it to ePub.

Amazon's pages that we read do not say how a Kindle sets the lines and text size of a PDF. For an older Kindle, see our guide to older Kindles, which uses ePub to MOBI.

A one-column novel: convert it, and fill in Title and Author

Our 19-page novel had one column, six chapters and one photo. After we pressed Convert File on the default settings, an ePub 3 file was ready in 6.7 seconds. By a plain count it held 9,493 words. We had numbered the novel's 120 paragraphs, and all 120 were in the ePub, in order. The six chapter titles became headings, and the photo stayed a picture.

Two things need your attention. The navigation list inside the file had one entry, “Start”, not six chapters. And with the Title (optional) and Author (optional) boxes empty, the book was named “in”, by “Unknown Author”. We typed “The Quiet Town” and “A. Writer”, and the file carried them. Both boxes sit under More settings.

Two columns and tables: keep the PDF

The four-page article had two columns, a table and, on every page, a running header and a page number. Column layout offers Auto-detect, Single column and Two columns, and it changed nothing. The article's three ePub files weighed 6,444, 6,444 and 6,445 bytes.

All 36 numbered paragraphs were in the article's ePub, but the order broke 15 times. The paragraph numbers ran 1, 6, 2, 7, 8, 3 and on. Lines from the two columns were joined with a vertical bar between them. The table was no longer a table: its cells ran on as lines of text among the paragraphs. The running header and page numbers were gone.

For a PDF like this, send the PDF itself. Send to Kindle and Kobo's eReaders both list it, and nothing in it gets mixed.

Wide margins: crop the PDF first

Crop PDF has a Trim white margins button. On the novel it took 0.4 seconds. A4 pages of 210 by 297 mm became pages of 177 by 259 mm, 0.84 times as wide. The text stayed text: page 1 still held 3,151 characters.

White margins use part of a small screen. We did not measure how much larger the text looks after trimming, because we did not open the file on a reader. To continue in another tool, PulsarKit may store a local copy of your input or result in this browser. It can remain after you close the tab or restart the browser. Clear PulsarKit’s site data to delete it.

A scan: run OCR in the language of the page

A scan is a picture of a page, so an ePub has no text to take from it until OCR reads the picture. With Run OCR left on Auto (only if scanned), PDF to ePub does that first. Our scan was one page of French, 318 words.

With OCR language on its default, English, the ePub matched 294 of the 318 words, or 92.5 percent. It misread accents: “trapéze” for “trapèze” and “plutét” for “plutôt”. With French chosen, it matched all 318 words and all 49 accented letters. Each run took about 13 seconds.

Set the language before you press Convert File: the menu starts on English and lists 15 languages. With Run OCR set to Never, the scan was refused, with a note to run OCR first and an Open OCR PDF button. The ePub also kept the page as a picture, so 469 kB came out as 462 kB. For a searchable PDF instead, use OCR PDF, which has its own language menu. Scanning a book covers that route.