Wise CSV Not Opening or Missing Transactions? Fix It

Official documentation reviewed: September 7, 2026

A Wise CSV can fail in two different places: the provider may have exported a narrower record set than you intended, or the spreadsheet may have interpreted the file incorrectly. Diagnose those separately. Editing cells before locating the failure can turn a recoverable import problem into an unreliable record.

Quick answer: Keep the original CSV untouched. Establish whether the missing or altered value exists in that original, check the export scope, then import a copy with explicit settings. Never invent rows or replace punctuation throughout the file to force it to open.

Wise CSV: locate the failure: Source — Expected row? — Correct scope?; Spreadsheet — Same value? — Same records?; Importer — Accepted rows? — Error report?; Fix — One stage — Retest sample — A parsing success does not prove complete transaction coverage.

Separate export scope from display errors

Symptom	First comparison	Next action
All data in one column	Raw headers versus imported columns	Review delimiter selection
Expected activity absent	Chosen record and filters	Generate the correct scope
Amount looks much larger	Original decimal characters	Review numeric interpretation
Rows appear twice	Source-file membership	Check overlapping imports
Identifier has changed	Original text versus stored cell	Re-import as text

Wise distinguishes currency-account statements from transfer lists. That distinction matters when the activity you expected used an external payment method. Return to the official statement scope and the transfer-list filters before concluding that an export is broken.

Recover a clean baseline

Make a folder containing the untouched download and a separate working copy. If you have already saved over the source, obtain a fresh export for the same scope. Record its currency, period and filter choices. Do not assume the latest file and an older file contain identical live activity simply because their filenames match.

Find one specific row or field that demonstrates the problem.
Inspect that value in a local plain-text viewer without editing it.
Compare it with the spreadsheet’s stored value, not just its visual formatting.
If it exists correctly in the original, focus on import settings.
If it is absent from the original, investigate the selected report and account view.
Write down the finding before changing anything.

Keep the test narrow. A single known reference and amount is easier to trace than an annual total that combines multiple currencies, fee treatments and overlapping files.

Import without silently changing values

In supported Excel versions, use Data → From Text/CSV and inspect the delimiter preview. Keep identifiers as text and review date and number types before loading. Microsoft explains that opening CSV directly uses default interpretation settings; its import tools provide greater control. Microsoft CSV import guidance.

Look for a clean separation of headers and values. Text descriptions may contain commas or quoted punctuation, so replacing every comma is not a safe repair. If characters look corrupted, inspect the import encoding rather than deleting the affected names. Preserve the full original file so you can restart.

An identifier displayed as scientific notation may merely be formatted differently, or it may already have lost precision. Compare the stored value with the original string. If digits or leading zeros changed, return to the source; guessing the missing characters cannot restore a reliable identifier.

Find the first point where rows disappear

Count records at each stage: downloaded source, imported worksheet and destination application. A text file’s physical line count is not always its transaction count because quoted fields can contain line breaks. Use a CSV-aware view for record counts, and account for the header. This is an original diagnostic method rather than a Wise-specific row guarantee.

Suppose an illustrative source has 120 transaction records, Excel has 120, and an accounting import accepts 118. That points to the last stage. Review rejected-row messages and the destination’s schema before downloading a different Wise report. If the source itself has only 118, compare scope and boundaries instead.

Remove worksheet filters in the working view before reporting missing rows. Then compare the date field used for grouping, the chosen currency and the exact period. Preserve pending explanations as notes rather than creating placeholder transactions in the accounting dataset.

Control joins and repeat imports

Add a source-filename column before combining monthly or currency files. A full-year export plus twelve monthly exports can duplicate the same records. Two genuine transactions can also share a date and amount, so those two fields alone do not establish duplication.

Compare available references, transaction type, currency and related details. Keep a review column stating why a row is included or excluded in the working result. Preserve related fee or conversion movements until you understand their role. Never delete source rows because they make a total inconvenient.

Before a full production import, validate a small supported test or preview in the destination software and check its duplicate-detection rules. Do not retry an entire import merely because the first screen looked unchanged; establish which records were accepted.

Prepare a useful support note

State the chosen export, date range, currency, spreadsheet or importer, exact symptom and the first stage where the mismatch appears. Use a fabricated sample to explain formatting in a public discussion. Send real references only through the provider’s authenticated support channel when necessary.

The private CSV diagnostic planner can turn the symptom into a checklist without receiving a file. Find other platform cases in CSV and Import Help, or review the record selection process in Transaction Histories. Keep your source, working copy and short change log until the discrepancy is resolved.

Trace overlapping Wise files: File A — Full-year rows — Original kept; File B — Monthly rows — Original kept; Overlap — Check references — Keep audit note; Output — Trace to source — No blind deletion — Same date and amount alone do not establish a duplicate.

Privacy: Financial exports can contain sensitive identifiers and counterparties. Troubleshoot locally and do not upload the source to an unfamiliar repair service.

Frequently asked questions

Should I save over the CSV after fixing it?

Use a new filename. Preserve the exact provider download as the baseline for later comparisons.

Can I remove duplicate-looking amounts?

Matching amounts do not prove duplication. Trace references, currencies, transaction types and source files before excluding anything.

Does a successful import prove the data is complete?

No. Compare coverage, rejected rows and control samples; successful parsing only establishes that the software accepted something.

Primary sources

Provider routes were checked against public documentation on September 7, 2026. Menu labels, available products and regional requirements can change. The examples and record-control checklists are original editorial guidance, not provider screenshots or a test inside your account.