Eight formats. Which one do you need?

What each file actually contains

Two transactions from a statement, written out by each of the four exports people ask about most. This is the shape of the file you get, not a description of it.

Sheet "Transactions"

Date        Description   Debit  Credit    Balance   Amount
2026-03-03  Card payment  42.60            1,866.31  -42.60
2026-03-05  Deposit              2,140.00  4,006.31  2,140.00

Date,Description,Debit,Credit,Balance,Amount
2026-03-03,Card payment,42.60,,1866.31,-42.60
2026-03-05,Deposit,,2140.00,4006.31,2140.00
2026-03-12,"Supplier, March",960.00,,2927.56,-960.00

OFXHEADER:100
DATA:OFXSGML
VERSION:102
ENCODING:USASCII

<INTU.BID>3000
<CURDEF>USD
<STMTTRN>
<TRNTYPE>DEBIT
<DTPOSTED>20260303120000
<TRNAMT>-42.60
<FITID>DEFF076FDD3FEEE4E36F84BD
<NAME>Card payment
</STMTTRN>

OFXHEADER:100
DATA:OFXSGML
VERSION:102
ENCODING:USASCII

<CURDEF>USD
<STMTTRN>
<TRNTYPE>CREDIT
<DTPOSTED>20260305120000
<TRNAMT>2140.00
<FITID>FEEE650FBC22434309EC69AC
<NAME>Deposit
</STMTTRN>
<LEDGERBAL>
<BALAMT>4006.31

QBO vs OFX vs QFX vs QIF

Four names, one family. OFX is the open standard; QBO and QFX are Intuit's branded variants of it; QIF is the older text format that came before all three.

Rule of thumb: QuickBooks → QBO. Quicken → QFX. Anything else that imports bank files → OFX first, QIF if OFX is refused. Read the longer explanation if you want the history.

Importing into QuickBooks Online

1. Open Transactions → Bank transactions.
2. Select the account you are importing into. If it is not connected to a feed, use Link account → Upload from file.
3. Choose Drag and drop or select files and pick the .qbo file you downloaded.
4. Pick the QuickBooks account in the Account name dropdown, then Continue.
5. If you are importing CSV instead, map the columns to QuickBooks' fields when prompted, then Continue.
6. Select the transactions to import, Continue, then confirm. They appear in the review queue.

Two limits worth knowing: QuickBooks expects the upload file to be in English, and CSV uploads have a small file-size ceiling, so a very long statement may need splitting, or use QBO instead. Menu names move between QuickBooks layouts; the steps above are the generic path. Last verified against Intuit's documentation: September 9, 2026.

The step by step, including what QuickBooks does with a date it does not recognize, is in importing a bank statement into QuickBooks Online.

Importing into Xero

1. From the Accounting menu, choose Bank accounts.
2. When Xero does not have the account, add it there and enter its details by hand. Choose Bank as the account type, or Other when the list does not offer Bank. Credit card is for card statements, and choosing it for a bank statement reverses every sign.
3. Find the account, open its menu and select Import a Statement.
4. Upload the .ofx file. Xero recommends OFX over CSV where possible; QIF also works.
5. If you use CSV, Xero wants income and expenses in a single amount column with expenses negative; our CSV export is already shaped that way. Map the columns when prompted.

Last verified by importing one of our files into Xero: September 11, 2026.

Xero's own PDF import covers a short list of countries. What the UK and Australia do instead covers the rest.

How the export behaves

• Bank identifiers in QBO. QuickBooks files carry a bank ID (INTU.BID). We write a generic one, which is what manual Web Connect imports accept. Some QuickBooks setups are configured to require a specific issuer instead: send us the bank name and we will add its identifier.
• Dates. Written in an unambiguous order in every format, so a DD/MM statement never turns into an MM/DD spreadsheet.
• Currency. The currency printed on the statement is detected and written into the export. One statement is treated as one currency, so a statement that mixes currencies in a single column comes back with those rows flagged rather than quietly summed across. Both number conventions, 1,234.56 and 1.234,56, are detected per statement rather than assumed.
• Descriptions. Exported exactly as printed, with no cleanup and no categorization. Merchant names are not guessed at and accounts are not assigned for you.
• Flagged rows. The reconciliation status travels with the data: a marked column in XLSX and CSV, and a per-row field in JSON, so a row that broke the chain is still identifiable after import.