How to import bank statements into Sage

Sage users get statement data in two ways: bank feeds for the connected present, and CSV imports for everything else — history, feed gaps, accounts without feed support, or a catch-up year. The import works well; the friction is entirely in getting a statement PDF into the CSV shape Sage expects. Here's the route.

Step 1: Statement PDF → CSV

Convert the statement with NoRekey and download CSV. Native PDFs and scanned paper both convert, and every conversion is verified against the statement's declared opening and closing balances first — Sage will happily import wrong numbers, so the proving belongs before the import, not after.

Step 2: Import — Sage Business Cloud Accounting

1. Go to Banking, open the account, and choose Import a bank statement — the option lives under the account's settings/connect menu.
2. Upload the CSV (Sage also accepts OFX/QIF here; OFX skips the column mapping).
3. Map the columns — Date, Description, and either a single signed Amount or separate Paid In / Paid Out columns — and confirm the date format (UK DD/MM/YYYY trips more imports than anything else).
4. Import; transactions land ready to match and categorise.

Step 2: Import — Sage 50

Sage 50 takes bank transactions via File → Import (or the Bank module's import option, edition-depending), with a CSV whose columns follow its bank-transaction template: transaction type, account reference, date, reference, details, and net amount. Build it once from your converted CSV, save it as the template, and every future statement is a paste job. Older Sage 50 installs can be particular about date format and header rows — if the import grumbles, check those two first.

Step 3: Reconcile the period

After importing, run Sage's bank reconciliation for the statement period: opening balance, closing balance, tick the rows. Because the conversion already proved the rows sum from the declared opening to the declared closing balance, this pass is confirmation rather than detective work — any discrepancy means a duplicate with feed data or an overlap at the period edges, not a misread.

The traps

• Feed overlap — importing a period the feed also delivered creates duplicates; trim the CSV to the gap, exactly as with Xero backfills.
• Summary rows — bank-generated CSVs sometimes append totals; delete anything that isn't a transaction before import. (Converted statements from NoRekey exclude balance and total rows already.)
• Signs — Sage's single-amount layout wants money out negative. If everything imports reversed, flip the sign column and re-import rather than editing rows one by one.