Convert a PDF's tables to Excel
To turn a PDF’s tables into Excel, open it in Convert from PDF and choose Excel workbook: each table becomes a sheet, every cell kept as the text it holds. CECO PDF refuses a PDF with a page that is not clearly a table, rather than guess, and names the pages.
Convert a PDF to Excel now — it opens Convert from PDF, with nothing to install; choose “Excel workbook” under “Format”.
- What it does: makes tables.xlsx, one sheet for each table, from the PDF’s own text or from a scan read the way OCR PDF reads one.
- What it refuses: a PDF with any page that is not clearly a table, since a workbook that left pages out would look complete when it is not.
- The price: Downloading costs nothing when nothing in the file you read was corrected here, even if it was merged, split or turned here first. A file made from a corrected document is priced as one, and so is one we can no longer trace back to the files it was made from. On the free plan, a conversion counts as half a correction of the day’s 10, once something is made.
The PDF you open is kept with your documents, and deleted after 90 days or whenever you delete it; deleting the export removes the workbook too. The privacy notice says what is kept, and why.
How do I convert a PDF to Excel?
- Open Convert from PDF and give it the PDF: “Drop a PDF here or choose one”. Up to 50 pages at a time.
- Choose “Excel workbook” under “Format”. Its note says it plainly: “Clearly tabular pages only. Cells are text; formulas and formatting are not recovered.”
- For a scan, choose the “Reading language”: English, Shqip, or “Find it for me”. It is used only where a page has to be read from its picture; a PDF with text of its own needs none.
- Press “Convert from PDF”. If a page is not a table, it says which, and nothing is made: “Some pages are not clearly tabular. Export their text instead.” Otherwise it says “Your PDF export is ready”, and “Download ZIP with record” gives tables.xlsx with record.json.
What counts as a table?
Rows of words that line up in columns. A table here is at least three rows that each have words in two or more columns, every cell lined up with the one above it at its left edge, its right edge or its centre. Ruled lines are neither needed nor looked for: the spaces between the columns are. So a price list or a timetable set out in columns is the kind of page that converts, ruled or not, and a letter, a page of running text or a form of labels and blanks is not. Only a table’s rows go into its sheet: a heading or a paragraph beside it on the same page is left out.
Why are my numbers text in Excel?
Because every cell is written as the text it holds, on purpose. A code with a leading zero keeps it, a date stays as it was printed, and a cell that starts with =, +, - or @ can never run as a formula when the workbook is opened. The price is that Excel does not add the figures up until you turn the column into numbers there. If you want numbers straight away, OCR PDF’s “Table (CSV)” writes a cell that is only a number as a plain number, so a spreadsheet adds it up; OCR PDF says how.
Can I convert a scanned PDF to Excel?
Yes. A scan is read the way OCR PDF reads one, in the language you choose, and a word read with too little confidence is left out and counted, never guessed: the cell where it stood is left empty, and the record says how many words each page left out. Check those cells against the scan. A photographed page is straightened first, by the slope of its lines, so a slight tilt does not fold one row into the next.
What you can and cannot do here
| What you want to do | Here |
|---|---|
| Convert a PDF whose every page is a table | Yes: a sheet for each table |
| Convert a PDF with tables on some pages and prose on others | No: it is refused, and the pages that are not tables are named. Take the table pages into a new PDF with Extract PDF pages first, or use OCR PDF’s “Table (CSV)”, which leaves the other pages out |
| Convert a scanned table | Yes: read as OCR reads it, with any word it is unsure of left out and counted |
| Keep a code’s leading zeros, and stop a cell running as a formula | Yes: every cell is kept as the text it holds |
| Get figures Excel adds up straight away | Not in the workbook: its cells are text. In OCR PDF’s “Table (CSV)”, a cell that is only a number is written as one |
| Keep colours, borders, merged cells or number formats | No: the cells and their text, in their rows and columns |
| Recover formulas | No: a PDF holds the figures a formula gave, not the formula |
| Convert more than 50 pages at a time | No: split the PDF first |
Questions people ask
Can I convert a PDF to Excel for free?
Downloading costs nothing when nothing in the file you read was corrected here, even if it was merged, split or turned here first. A file made from a corrected document is priced as one, and so is one we can no longer trace back to the files it was made from. On the free plan, a conversion counts as half a correction of the day’s 10, and only once a workbook is made. The plans are on the pricing page.
Why does it say my pages are not clearly tabular?
Because the pages it names hold words that do not line up in columns: running text, a form, or a table too short to tell from chance. Nothing is made rather than a workbook with those pages missing. “Text with layout” keeps such a page’s columns in plain text, and Convert from PDF lists every format.
Will the Excel file be editable?
Yes. Every cell is ordinary text you can change, sort and copy, and none of them is a formula. Turn a column into numbers in Excel to calculate with it.
Can I convert a PDF to Excel without losing the formatting?
No. The rows, the columns and the text in each cell come across; colours, borders, merged cells, fonts and number formats do not, and the record says so.
How are the sheets named?
One sheet for each table, named by the page it is on and its place in the workbook, such as “Page 1 table 1”.
Can I turn the spreadsheet back into a PDF?
Yes: Excel to PDF makes a PDF of an Excel workbook, and says what it keeps.
What does the record of an Excel export say?
How many tables and rows each page gave, how many words a scan left out, and whether the words came from the PDF’s own text or were read from its picture, with the PDF’s fingerprint and the workbook’s. None of the words themselves.
Related pages
- Convert from PDF, for text, Word, pictures or an archival copy
- OCR PDF, for a table as CSV, and a searchable copy of a scan
- PDF to text, for the words of pages that are not tables
- Extract PDF pages, to take the pages with tables into a PDF of their own
- Excel to PDF, to turn a workbook into a PDF
Last updated