How to Extract Tables from PDF to Excel Spreadsheets

A table in a PDF only looks like a table. The file stores each number and word as a separate piece of text at a position on the page. It has no rows, no columns and no cells. An extraction tool rebuilds the grid from those positions: pieces at the same height become a row, and a wide horizontal gap marks the start of a new column. This works well for tidy tables and less well for crowded ones, so the real job is knowing how to check and repair the result.

Worked example: a three page bank statement

Say you have a statement with the columns Date, Description, Debit, Credit and Balance, and you want to add up your spending by month.

1. Convert the PDF with the PDF to Excel tool and open the .xlsx file. Each page of the statement is on its own sheet.
2. Copy the rows from the second and third sheet underneath the rows of the first, so everything is in one list. Delete the header row that repeats on each page.
3. Look at the Description column. Long descriptions that wrapped onto a second line in the PDF now sit in a row of their own with no date and no amount. Move that text up into the row it belongs to and delete the leftover row.
4. Click a cell in the Debit column. If the number is aligned to the left, Excel is treating it as text, often because of a thousands separator or a currency sign. Select the column and use Find and Replace to remove the separator or the sign, then set the cell format to Number.
5. Now the important part: add up the Debit column and compare your total with the total printed on the statement. Do the same for Credit. If both match, the extraction is complete. If not, a row was split or lost, and you can find it by comparing page by page.

That last comparison takes a minute and is the only way to know the data is right. Never skip it for anything involving money.

Common mistakes

• Trusting the numbers without a control total.
• Forgetting that debits and credits may land in the same column when the PDF places them close together. Check a few rows against the original.
• Leaving numbers as text. Sums over text cells show zero, which looks like a broken formula.
• Overlooking dates. A date like 03/04/2026 can be read as the third of April or the fourth of March, depending on the regional settings of Excel. Check one date you know.
• Converting a scanned statement and expecting perfect digits. OCR can misread a 5 as a 6. With scans, check every amount.

Troubleshooting

• Everything is in one column: the gaps between the columns of the PDF are narrow. Use Text to Columns on the Data tab in Excel and split by spaces, then repair the description column.
• One column is split into two: a wide gap inside a cell was taken for a column border. Join the two parts with a formula such as =A2&" "&B2, or simply retype the few cells affected.
• Rows are out of order: this is rare, and happens with rotated pages. Rotate the page in a PDF viewer, save and convert again.

When not to use this method

Look for a better source first. Many banks and accounting systems let you download the same data as a CSV or Excel file, and that will always be cleaner than anything pulled out of a PDF. For a table of five or six rows, typing it is faster than converting and checking. And for tables with merged cells, several lines of text per cell or nested headings, expect to spend real time on repairs.

Checklist

• I looked for a CSV or Excel download before converting.
• All pages are combined into one sheet and repeated headers are removed.
• Wrapped descriptions are merged back into their rows.
• Number columns are real numbers, not text.
• My totals match the totals printed in the PDF.