Bank Statement PDF to Excel [Free Tool 2026] - 30 Seconds

Most banks will hand you a PDF and nothing else. If you want to add up a category, check a year of subscriptions, or hand a bookkeeper something they can work with, the PDF is the wrong shape and retyping it is a bad evening.

Our bank statement converter turns that PDF into a spreadsheet. It is free for 30 pages a month, and it does not need to know which bank you are with.

Rateweb analysis, 2026

We ran our own parser against representative statement layouts from four countries to find where converters go wrong. The failure we found is one almost nobody checks for:

Statement style	Date read as	Correct?
US 03/04/2026	4 March	yes, after our fix
US 12/25/2026	25 December	yes
UK 03/04/2026	3 April	yes
Canada 2026-03-04	4 March	yes
Australia 03/04/2026	3 April	yes

Before the fix, a US statement produced a MIXTURE of correct and incorrect dates in the same file: 03/04 was read day-first and wrong, while 12/25 fell through to a US reading and was right. Nothing marked which was which. Any converter that assumes one date order without checking the document will do the same to your data, silently.

Method: parser output compared against known-correct dates for each layout, covered by automated test. Analysis by Rateweb.

First, check whether you need this at all

Worth saying before anything else, because it costs us a conversion and saves you a step.

If your bank offers a CSV, OFX or QIF export, use that instead. Several do, and it is usually buried a click or two away from the statement download — look for "export", "download transactions", or a format dropdown next to the date range.

A native export is better than any conversion of a PDF, for a simple reason: it is the data the bank holds, rather than a reconstruction of a document the bank printed. Nothing has to be inferred, so nothing can be inferred wrongly.

Convert the PDF when there is no export available, when the export does not go back far enough — banks frequently cap CSV history at a few months while offering PDFs for years — or when someone has handed you statements rather than access to the account. Those cases are common, which is why this tool exists. But if the CSV is sitting there, take the CSV.

Why it works across banks

Most converters are built per bank: someone maps one layout, and the tool breaks the moment the bank redesigns its statement or you switch accounts.

This one reads structurally instead. On each line it looks for a leading date, then the money amounts on that line. That pattern holds across most retail statements regardless of who printed it, because a statement is fundamentally a dated list of amounts.

The consequence is that an unusual or foreign bank is not automatically a problem, and a layout change does not silently break anything.

Debits and credits come from the balance, not the layout

This is the part most converters get wrong.

When a PDF is converted to text, the column positions are lost. A "-45.00" in the debit column and a "45.00" in the credit column can end up looking identical. Tools that guess from the printed minus sign get it wrong on every bank that uses separate debit and credit columns instead.

We derive the sign from the running balance instead. If the balance fell by 45.00 on that line, it was money out, whatever the layout was doing. Where a balance column exists, this is reliable in a way that reading the sign is not. Where there is no balance column, we fall back to the printed sign.

Dates are read in the order your statement actually uses

03/04/2026 is 3 April in South Africa, the UK and Australia. It is 4 March in the United States and Canada. Both readings parse cleanly, so a single date cannot tell you which one is meant.

Getting this wrong is worse than it sounds, because it fails silently and inconsistently: a tool that assumes day-first will read 03/04 as 3 April (wrong for a US statement) but read 12/25 correctly, because there is no 25th month to fall back on. You end up with a spreadsheet where some dates are right and some are wrong, and nothing marks which is which.

We decide the order once for the whole statement, using evidence in the document. A number above 12 in the first position cannot be a month, so the statement is day-first. A number above 12 in the second position cannot be a month, so it is month-first. Whichever the file proves, we use.

Tested against representative layouts:

If a statement is genuinely ambiguous — every date has both numbers at 12 or below — we default to day-first. Check a few rows against the original if that is your situation.

What we keep

We record that a conversion happened: your user id, the number of pages, and whether the source was a PDF or a CSV. That is what the monthly page count is counted from.

We do not keep the statement or the transactions in that record. The file is uploaded and parsed on our servers — anyone telling you a PDF converter works without the file leaving your device is describing a different kind of tool — so treat it as you would any upload, and redact anything you would rather not send if you are dealing with a particularly sensitive account.

If your statement does not parse

Three things account for most failures:

1. A scanned or photographed statement. If the PDF is an image of a page rather than text, there is nothing to read. Download the original from your bank's site rather than scanning a printout.
2. A password-protected file. Many banks send statements encrypted. We handle the common case where the file opens without prompting you, and you can supply a password otherwise.
3. A layout with no leading date per line — some credit-card statements group transactions under a date heading instead of repeating it. Those are the hardest case and may come through incomplete.

If you get partial results, check the row count against the statement before relying on it. A converter that quietly drops rows is worse than one that fails loudly, so it is worth thirty seconds.

What to do with the spreadsheet

Converting is the means, not the end. Three things are worth doing with the output, roughly in order of how much they pay.

Find what repeats. The recurring charges are where the money usually turns out to be — the forgotten subscription, and more often the one that has quietly risen. Find every recurring bill does this against the same import.

Turn the category totals into a budget's actual column. Most budgets fail because they only ever record the plan, and nobody wants to type a month of spending. Reading it off a converted statement takes minutes. The budget planner holds planned and actual side by side.

Hand it to whoever asked. A bookkeeper or accountant billing by the hour should not be retyping a PDF. A clean spreadsheet is the cheapest thing you can give them.

One correction before you read any total: take out transfers between your own accounts, and payments to your own credit card. They are not spending, and left in they inflate income and expenditure at the same time — so the two errors hide each other and the numbers look consistent while being wrong.

How does this affect YOUR Money OS?

A year of categorised transactions is the input your budget and net worth tracking need. Most people are guessing at both until they have it.

Check my free OS score

FAQ

Can I convert a bank statement PDF to Excel for free? Yes, 30 pages a month here, with no account needed to see the preview.

Will it work with my bank? It reads line-based PDF statements structurally rather than per bank, so an unusual or foreign bank is not automatically a problem. Scanned or photographed statements will not work, because there is no text to read.

Why are some transactions uncategorised? Deliberately. Several chains share a name with an ordinary English word, and matching them would fill in more categories and be wrong some of the time. A blank category is visible; a wrong one is not.

Is my statement stored? The conversion record keeps your user id, the page count and the source format. Not the statement, and not the transactions.

Does it handle US date formats? Yes. Date order is detected per statement from evidence in the document rather than assumed.

Should I use a CSV export from my bank instead? Yes, if one is available. A native export is the data the bank holds rather than a reconstruction of a printed document, so nothing has to be inferred. Convert the PDF when there is no export, when the export does not go back far enough, or when someone handed you statements rather than account access.

Why do my spreadsheet totals look too high? Almost always transfers between your own accounts, and payments to your own credit card. Both appear as money leaving, and often as money arriving on the way back. Remove them before reading anything.

Can I convert several months at once? Yes, and it is the sensible way to do it — a year in one pass, named so the files sort themselves. See how to organise 12 months of statements.

Bank-specific pages

Same tool, named for the bank you are with:

• UK — Barclays, HSBC UK, Lloyds, NatWest, Monzo
• US — Chase, Bank of America, Wells Fargo
• Canada — RBC, TD Canada Trust
• Australia — Commonwealth Bank, ANZ
• South Africa — FNB, Capitec, Absa, Nedbank, Standard Bank