Trial Balance Does Not Match After a QuickBooks Migration

When a trial balance does not match after a QuickBooks Desktop to Online migration, rule out a false alarm before investigating anything. Confirm both reports use the same accounting method and the same date, and confirm neither is broken out into multiple periods. Those two mistakes produce dozens of fake differences and account for most reported mismatches.

What should I rule out first?

Two settings, in this order. Both are quick, and both invalidate everything downstream if they are wrong.

Check	Why it matters	Where to look
Accounting method	Cash versus accrual changes nearly every account	Top of the Desktop report, bottom of the Online export
Column layout	One Debit/Credit pair per month reads as the wrong period	The report header row
As-of date	An obvious one, easily wrong	The report subtitle
Zero-balance rows	Suppressed rows hide missing accounts	Report customisation

The accounting method trap is worth dwelling on. QuickBooks Online prints Accrual Basis as the very last row of an exported file, after all the data, while Desktop puts it near the top. Check both ends of both files.

Why would the column layout make it wrong rather than obviously broken?

Because a multi-period report still adds up.

A Trial Balance exported with columns broken out by month carries one Debit/Credit pair per period. Read the first pair and you are looking at the earliest month while the heading says "As of" your migration date. The report's own total row is broken out the same way, so the columns still reconcile internally. Everything looks consistent and everything is the wrong period.

Set "Display columns by" to Total before exporting.

What is the actual diagnostic order?

Once the false alarms are ruled out, work from broad to narrow. Each step should reduce the search by roughly an order of magnitude.

1. Compare account by account, not by total — both files balance regardless
2. Note whether one difference explains another, such as inventory against Opening Balance Equity
3. Separate expected artifacts from real problems before chasing anything
4. For A/R and A/P, go to the ageing detail — the control account cannot name an invoice
5. For inventory, go to the Inventory Valuation Summary
6. For anything still unexplained, compare the Account Listing for a changed type

What counts as an expected artifact?

Finding	Expected?
Opening Balance Equity has a balance	Yes — it is the migration's plug account
Sales tax account renamed to an agency	Yes — Online names them that way
Inventory value changed	Common, but always worth explaining
An account missing entirely	No
An account with a different type	No
A difference under a dollar	Rounding — set a materiality threshold and move on

Sorting into these two buckets first is what keeps the investigation finite. Chasing a $0.01 rounding difference with the same energy as a missing $2,300 invoice is how a morning disappears.

How do I know I have finished?

When every remaining difference has a written explanation, and the only balance left in Opening Balance Equity is genuine opening equity.

Not when the totals agree. They agreed at the start.

What these traps look like in a real file

The multi-period trap is not hypothetical. A real QuickBooks Trial Balance export we parsed carried 36 Debit/Credit column pairs — three full years of months — under a heading that read "As of December 31, 2025". Read positionally, the first pair returns January 2023 figures. Checking reads 143,851.32 instead of its actual December-2025 balance of 410,900.99, a difference of over $267,000 on one account, with no indication anything is wrong.

The same export printed Accrual Basis on its last line, line 93 of 94, after all the data. Anyone checking the top of the file concludes the basis is unknown.

A correctly exported single-period version of that same file contains 83 accounts that reconcile exactly to the report's own total row. That reconciliation — adding up what you extracted and checking it against the printed total — is the cheapest check available and it catches every one of these traps at once.

Quick reference

Question	Answer
First thing to check?	Accounting method, at both ends of both files
Second?	Whether the report is broken out by period
Do totals agreeing prove anything?	No — both files always balance
Where do A/R differences get resolved?	The ageing detail, never the trial balance
When are you done?	Every difference explained, Opening Balance Equity at zero

The free check runs the false-alarm tests first and refuses to compare mismatched reports rather than showing you differences that are not real.