Bank statement to Excel, without uploading it anywhere

Convert the statement now

Before you go looking for a converter at all: check your online banking. Many banks offer a transaction export as CSV or Excel, and that is the bank’s own data with nothing reconstructed. Convert the PDF when the export does not exist, does not reach back far enough, or when the PDF is all you were sent.

Why a bank statement should not go to an upload converter

A statement is the most revealing document most people own. It carries your account number and IBAN, your balance, your salary and your employer, your rent and your landlord, and every shop, doctor and subscription you paid that month. An online converter that uploads it hands all of that to a server you know nothing about, often under a free tier that keeps files “for 1 hour”.

The converter on this page runs in your browser. The file is read in your tab and the spreadsheet is written there; you can test that by converting once (the first file fetches the engine, which your browser then keeps), turning Wi-Fi off and converting again. The wider story is in PDF privacy and protection.

The steps, with a real result

1. Drop the PDF on the tool above. A statement downloaded from online banking has real text in it, which is what the converter reads.
2. Pick the sheets. The default is “One per page”. For a statement of several pages, choose “All on one”, so the transactions run in 1 list (pages are separated by an empty row).
3. Keep “Excel (.xlsx)” as the format unless you are importing into accounting software; for CSV, pick the separator that matches your Excel (below).
4. Convert and open the file. The result card says how many rows it wrote and how many cells became numbers.

We ran an invented October statement for Jana Čermák in Linz through the converter: an opening balance, 6 transactions with European decimal commas, and a closing balance. The card said “Put 12 rows into a spreadsheet, with 14 numbers as numbers you can add up”. The transaction part of the sheet looked like this:

The statement printed “-1.250,00”; the sheet holds the number -1250, which Excel shows in whatever number format you give the column. The header lines above row 4 (title, account holder, account number) come along as single cells, so nothing on the page is lost. The result was the same whether the statement drew lines between its columns or only left gaps.

Check it in 30 seconds: sum against the balance

Never trust a converted statement until it adds up. In the sheet above:

1. In an empty cell, type =SUM(C5:C12). It gives 2001.31.
2. In the next cell, type =D12-D5, the closing balance minus the opening balance. It also gives 2001.31.
3. If the 2 match, every amount arrived, with the right sign. If they do not, the difference is the size of the missing or misread amount, which tells you which row to look at.

This check catches every kind of miss at once: a skipped row, an amount in the wrong column, a minus sign lost. Do it on the first statement from each bank; after that the layout is known and you can spot-check.

Several months in 1 sheet

1. Convert each month’s statement to its own .xlsx, with “All on one”.
2. Make a new workbook with the same 4 columns, plus a fifth called Statement.
3. Paste each month’s transaction rows under the last, without the header rows and without the opening and closing balance rows, and fill the Statement column with the month (2026-10, 2026-11).
4. Run the sum check per month before you paste, not on the combined sheet: a miss is far easier to find in 1 month than in 12.

The Statement column lets you filter back to a single month later and keeps a trail to the source PDF.

The traps, and what fixes each

• Dates arrive as text. The converter keeps 01.10.2026 as it is written, on purpose: a US-set Excel would otherwise read it as 10 January. To sort by date, select column A, then Data, Text to Columns, Next, Next, choose Date and DMY, Finish.
• Account numbers and IDs keep their leading zeros. A value like 0012345678 stays text, so Excel does not shorten it to 12345678, and a long card or reference number is not turned into scientific notation. PDF to Word, Excel and text explains how the converter tells numbers from IDs.
• A description that wraps onto 2 lines becomes 2 rows. We tested it: “Rent October,” stayed in the transaction row with its amount, and “Graz Lettings” landed in a row of its own in column A. The sum is not affected, because the amount stays in its row, but delete or merge those half rows before sorting.
• CSV in a European Excel. An Excel set to German, French or Polish expects semicolons between fields and a decimal comma, so a comma CSV double-clicked there lands entirely in column A. When you pick CSV, set Separator to “Semicolon, 1190,50”: the fields are split by semicolons and -64.19 is written -64,19, so that Excel opens it in columns with the amounts as numbers. “Comma, 1190.50”, the default, suits Excel in English, Google Sheets and most software. The .xlsx needs neither: it opens correctly in every language. For accounting software, CSV is usually the format the import asks for.

When the statement is a scan or a photo

A statement you scanned, photographed or received as a picture has no text in it, only pixels. The converter detects this and stops with a pointer: “There is no text in this file to put in a spreadsheet: it is probably scanned. Run OCR PDF first, then convert it.”

We tested that route too: the same statement rendered as a 200 dpi greyscale scan, run through OCR PDF (94 % average confidence), then converted. Every amount and balance came through as the right number. The difference: the date and the description landed together in column A (“02.10.2026 Rent October, Graz Lettings”), because a scan’s column lines are part of the picture, not drawn lines the converter can use. Two formulas split them again: =LEFT(A6,10) for the date and =MID(A6,12,200) for the description. The sum check matters even more here, because OCR can misread a digit. The full OCR route is in Scanned PDF to searchable.

The honest part

Statement layouts differ from bank to bank, and the converter works out columns from where the text sits on the page. Most statements are a clean grid and convert like the example above. Statements with 2 tables side by side, a running balance printed in the description, or amounts split into separate debit and credit columns convert too, but need a look before you rely on them; the result card warns about this every time. It does not keep colours, fonts or cell borders, and it does not know which column is the amount: you tell Excel that with your formulas. And if your bank offers a CSV export, that file beats any conversion, ours included.