How to Read a Bank Statement
A statement is four things stacked together: who the account belongs to, what it was worth at the start, everything that happened, and what it was worth at the end. Once you can see those four, the rest is labelling.
Published 9 August 2026 · 8 min read
The short answer
Read it in this order: the period at the top, the opening balance, the transaction table, the closing balance. If the opening balance plus everything in the middle does not land exactly on the closing balance, something is missing, and that check is the only part most people skip.
Statements look more complicated than they are, mostly because banks use different words for the same things and one piece of the vocabulary is genuinely backwards. This goes through one top to bottom.
The header: whose account, and when
The top of every statement identifies the account and the window it covers. Four things live there.
- The account holder's name and address. This is why a statement is accepted as proof of address: it ties a name to a place on a document a third party produced.
- The account number, usually partly masked, and a sort code, routing number or IBAN depending on the country.
- The statement period, the start and end dates. Everything inside belongs to this window and nothing outside it does.
- The statement date, which is when it was produced. It is often a day or two after the period ends, and it is not the same thing.
The statement period matters more than people expect. If you are reconciling and your figures are out, the first thing to check is whether the period you exported matches the period you think you exported. An off-by-one day at either end explains a great many mysterious discrepancies, and it is much easier to check than to deduce.
Opening and closing balances
Under or beside the header there will be a small summary: what the account held at the start of the period, and what it held at the end. Names vary: opening and closing, brought forward and carried forward, previous and current, starting and ending.
These two figures are the most useful numbers on the whole document, because together with the transactions they make the statement checkable:
Opening balance + every credit − every debit should equal the closing balance, exactly. If it does not, something is missing from the middle.
That identity is what makes it possible to prove a statement was read completely rather than nearly completely, which is the whole game when a page break or a bad export has quietly eaten a row.
Available balance and current balance are not the same
Online banking usually shows two figures and they frequently disagree.
| Figure | What it means |
|---|---|
| Current, ledger or book balance | What has actually settled. This is the figure a statement shows. |
| Available balance | What you can spend right now: the settled figure adjusted for pending card authorisations, uncleared deposits and any overdraft. |
A card payment made this morning has reduced the available balance and not the current one. A cheque paid in may have raised the current balance while the funds are not yet available. Neither is wrong; they answer different questions.
For reconciliation, the settled figure is the one that matters: pending items are not on the statement and should not be in the books either.
The transaction table
The body of the statement is a table with one row per transaction. Columns vary by bank, but they are drawn from a small set.
| Column | What it holds |
|---|---|
| Date | Usually the posting date: when it settled, not when you spent it |
| Description | Merchant, payee or a reference, often abbreviated and truncated |
| Type | A code for the method: card, direct debit, transfer, cheque |
| Debit / paid out | Money leaving the account |
| Credit / paid in | Money arriving |
| Amount | Some banks use one signed column instead of the two above |
| Balance | The running total after this row |
The single-column and two-column layouts are the main structural difference between statements. A US statement usually has one signed Amount column; a UK or European one more often splits paid in and paid out. Both describe the same thing, and a tool reading either has to work out which it is looking at.
Transaction date versus posting date
Card transactions carry two dates: when you made the purchase, and when the bank settled it. Statements generally sort by the posting date, which is why a Friday-night restaurant bill can appear on Monday.
The consequence worth knowing is at period boundaries. A purchase made on the 30th can post on the 2nd and land in the following statement, so a receipt and a statement line can sit in different months without either being wrong.
This is a routine source of apparent discrepancies and almost never an error. Before hunting for a missing transaction at the end of a period, check the start of the next one.
Why credits and debits look backwards
This trips up nearly everyone who moves between banking and bookkeeping, and it is worth understanding once properly.
On your statement, money arriving is a credit and money leaving is a debit. That is the opposite of what a bookkeeper does with the same event, where receiving money into a bank account is a debit to that account.
Both are correct, because the statement is written from the bank's point of view. Your deposit is money the bank owes you, a liability on their books, and liabilities increase with credits. Your account is an asset on your books, and assets increase with debits. Same event, two sets of books, opposite words.
The practical version: on a statement, credit means in and debit means out. Do not try to reconcile the terminology with double-entry while you are reading it.
The running balance column
Where a statement has a balance column, each row shows what the account held immediately after that transaction. It is more useful than it looks, for one reason: it makes every single row independently checkable.
Take any row's balance, subtract the row above it, and the difference should be that row's amount. Where it is not, either a transaction is missing between them or the rows are not in the order you assumed: some exports run newest first.
That row-by-row check is stronger than comparing opening and closing totals, because it localises the problem instead of only proving one exists. If forty rows step correctly and the forty-first does not, you know exactly where to look.
Why the descriptions are so cryptic
Statement descriptions pass through several systems with fixed field widths, and each one truncates. The OFX format that statement files are built on limits its main name field to 32 characters, which is why the same shop can appear under two or three slightly different names in one account.
On top of that, many merchants bill under a trading entity or a payment processor rather than the name over the door, so an unfamiliar payee is more often a familiar one in disguise than a problem.
The codes themselves, BACS, CHAPS, POS, ACH, NSF and the rest, are covered in bank statement abbreviations.
Fees, interest and the small print
At or near the end there is usually a section covering charges. Account maintenance fees, overdraft or unarranged borrowing charges, foreign transaction fees, interest paid to you, interest charged to you.
These are real transactions and they belong in the books like anything else, but they are easy to skip because they arrive on their own rather than being something you did. On a business account they are also deductible, which makes missing them a small ongoing cost rather than a tidiness problem.
Getting hold of older statements
Online access to past statements is commonly one to seven years depending on the bank and the account type. Older records generally still exist but have to be requested, sometimes for a fee and usually with a wait measured in weeks.
The practical advice is dull and worth following: download anything you might need for tax, an audit or a loan application while it is still in the app, rather than assuming it will be there when you go looking. This applies doubly to an account you are about to close, where access frequently ends with the account.
Checking a statement adds up
If you are reading a statement to enter it somewhere, do this before you rely on it:
- Confirm the statement period is the one you meant to pull.
- Total the money in, total the money out.
- Add the difference to the opening balance and check you land exactly on the closing balance.
- If you are out, use the running balance column to find where: the first row whose step does not match its amount is the place to look.
In a spreadsheet that is two formulas and takes a minute, and it is worth doing by hand at least once so you know what the check is actually proving.
If the statement came as an export you need in a spreadsheet or in accounting software, converting it does that arithmetic on the way through and names the exact size of any gap. It runs in your browser and needs no account.
Convert a statementFrequently asked questions
What does the statement period mean? The window of time the statement covers, given as a start and end date. Every transaction inside belongs to that window. It is not the same as the statement date, which is when the document was produced, usually a day or two later.
Why is my available balance different from my current balance? The current or ledger balance is what has settled. The available balance adjusts that for pending card authorisations, deposits that have not cleared and any overdraft facility. A card payment made today has usually moved one and not the other.
Why does a credit mean money coming in? Because the statement is written from the bank's side. Your deposit is money the bank owes you, and on their books that liability increases with a credit. In your own bookkeeping the same deposit is a debit to your bank account. Both are right; they are different sets of books.
How long do banks keep statements? Online access is commonly one to seven years depending on the bank and account, and older records generally exist but must be requested. Download anything you may need for tax or an audit while it is still in the app rather than assuming it will be there later.
Can a bank statement be used as proof of address? Usually, when it is recent, often within three months, and shows your name and address as the bank holds them. Requirements differ by who is asking, and some will not accept a self-printed copy.
What if my transactions are in reverse order? That is common in exports, which frequently run newest first while paper statements run oldest first. It changes nothing about the figures, but it does invert the running balance check: each row's balance relates to the row below it rather than above.