Bank statement PDF to Excel: getting a multi-page statement into one clean table

Excel can read tables straight from a text-layer PDF. A one-page statement is easy. A statement that runs over several pages is where people get stuck, because Excel hands you several separate tables. This page shows the free fix, then the other problems you will meet.

1. Basic import

In Excel: Data > Get Data > From File > From PDF, pick the file.
In the Navigator, pick the table(s) you want. Choose Load, or Transform Data to clean first.

Source: Microsoft's Power Query PDF connector page (read 2026-10-04): "In Navigator, select the file information you want, then either select Load to load the data or Transform Data to continue transforming the data in Power Query Editor." The connector's listed products include Excel.

It only reads PDFs that have a text layer. A common summary, from a search result on 2026-10-04, is that a scan with no OCR returns no tables. Test by selecting text in the PDF first.

2. The multi-page problem

A user on Microsoft Q&A describes it: "the one table that displays over three pages in the PDF is identified as three separate tables in the Get Data Navigator screen ... only the first table contains the original column titles, and the other two tables display 'Column1, Column2, etc.'" (Microsoft Q&A thread, read 2026-10-04.)

The accepted answer in that thread, so you do not rename columns by hand:

Open Table 1 in Power Query. On the Transform tab, open the arrow under Use First Row as Headers and choose Use Headers as First Row.
On the Home tab, choose Append Queries and add Table 2, Table 3 and so on. Now all of them have the same generic column names, so they stack.
On the combined table, Transform > Use First Row as Headers.

The original headers repeated on later pages will now appear as rows in the data. Filter those out (for example filter the Date column to remove the text "Date").

3. Wrapped (multi-line) descriptions

Microsoft's connector page says: "In cases where multi-line rows aren't properly identified, you might be able to clean up the data using UI operations or custom M code. For example, you could copy misaligned data to adjacent rows using Table.FillDown, or group and combine adjacent rows using Table.Group."

In plain terms: a second description line shows up as its own row with no date and no amount. Either fill the date down and then group, or join the description onto the row above before deleting the stray row.

4. Large statements

Microsoft's page suggests, for large PDFs that time out, selecting pages one at a time with the StartPage and EndPage options of the Pdf.Tables function, and, if the PDF is one huge table, disabling the MultiPageTables option. This needs editing the formula bar or Advanced Editor.

5. Finish the table

Convert the Date column to a real date type, and check the day/month order on a row where the day is above 12. Regional settings change how text is read.
If money in and money out are separate columns, make a signed Amount: =Credit-Debit.
Remove totals, "continued" lines and opening/closing balance rows from the transaction list.
Check: opening balance + money in - money out = closing balance. If it differs, you lost, duplicated or mis-signed a row; the multi-page append is a likely cause.

If you then need the data in an accounting package, see the QuickBooks and Xero pages for the CSV formats.

6. If you want a script instead

I sell a one-time $29 Python script that reads text-layer PDF statements and writes CSV, which Excel opens. It is a script you run on your own computer, not a website. You need Python installed and one install command (pip install pdfplumber). It does not read scanned or password-protected PDFs, and I have not tested it on a real bank's layout or on a multi-page statement end to end. Bank Statement PDF to CSV Converter on Gumroad. For one-off statements, Excel's own import above is free and probably enough.