Intuit’s free migration tool moves your lists and most of your transaction history from QuickBooks Desktop to QuickBooks Online, but it does not move payroll detail beyond the current year, your audit trail, reconciliation reports, inventory assembly structures, price levels, or more than one A/R and one A/P account. Re-running a migration completely overwrites the target QBO company, destroying anything entered since the first attempt. And despite the panic, QuickBooks Desktop itself keeps working after its support date passes; only its connected services stop. For some businesses, not migrating is the correct decision.
We wrote this because the search results for “QuickBooks Desktop to Online migration” are Intuit’s own help articles, which cover the how but structurally cannot cover the whether, and a handful of Reddit threads from people who found out the hard way. We help small businesses run these migrations, and a meaningful share of our assessments end with “don’t.” If you are an accountant or bookkeeper advising a client, this article is written to be forwarded.
What actually happened, and what actually stopped
The urgency most owners feel comes from Intuit’s rolling three-year support policy. QuickBooks Desktop 2021 lost support on May 31, 2024. The 2022 versions followed on May 31, 2025, and the 2023 versions on May 31, 2026. Desktop 2024 and Enterprise 24.0 are supported through September 30, 2027.
When your version crosses its date, a specific list of things stops working: bank feeds, including transaction upload; payroll, where Basic, Standard, Enhanced and Assisted all deactivate; payments processing; live support; security updates; Accountant’s Copy transfers; multi-currency features; in-product email; and Shipping Manager.
Here is what does not stop: the software. QuickBooks Desktop keeps opening. Bookkeeping, invoicing, reporting and printing continue indefinitely. This is the single most important fact in the whole discussion, because most of the migration panic we see is built on the belief that the program will brick itself on the cutoff date. It will not.
It is also worth being precise about the sales side. Intuit stopped selling Pro Plus, Premier Plus, Mac Plus and Desktop Enhanced Payroll to new US subscribers on September 30, 2024. But existing subscribers can renew indefinitely, with continued updates and support, and Enterprise is still sold to new customers. There is no announced full sunset of QuickBooks Desktop. Fourlane, an Elite Intuit Solution Provider that cites over one million active Desktop users, put it plainly in 2026: Desktop is “still strong, still improving, and actively supported.”
What is true is that staying keeps getting more expensive. On October 1, 2025, Accountant went from $1,199 to $1,799 and the Premier Plus per-seat price went from $300 to $465. On February 1, 2026, Pro Plus and Mac Plus went from $999 to $1,149, Premier Plus base from $1,399 to $1,609, and Enterprise Silver from $1,703 to $1,873. The pressure to move is real. It is financial and gradual, not a cliff.
What does not migrate: the full table
This is the list Intuit confirms in its own documentation, gathered in one place. Print it, walk your file against it, and you will know most of your migration risk before you touch anything.
| Data in QuickBooks Desktop | What happens in QuickBooks Online | Practical consequence |
|---|---|---|
| Payroll detail beyond the current year | Prior-year paychecks convert as regular checks with no payroll item breakdown; current-year payroll arrives as lump sums | Historical payroll reporting is gone; year-to-date figures must be re-entered in QBO Payroll by hand |
| QuickBooks Time timesheet data and settings | Does not convert | Time history and configuration must be rebuilt or archived separately |
| Audit trail | Does not convert | Your record of who changed what starts over on migration day; keep the Desktop file as evidence |
| Past reconciliation reports | Do not convert; cleared status does carry on individual transactions | The first QBO reconciliation must be set up carefully, and its opening balance is usually wrong out of the box |
| Multiple A/R and A/P accounts | QBO allows only one default A/R and one default A/P account | Files structured around multiple receivable or payable accounts need re-architecture before migration, not after |
| Price levels | Do not exist in QBO; billing rate levels and sales reps also do not convert | Customer-specific pricing must be rebuilt with QBO price rules or handled manually |
| Units of measure and manufacturer’s part numbers | Do not convert | Item records lose data that purchasing and fulfillment may depend on |
| Inventory assemblies | Become single inventory items on sales transactions; the build structure is effectively lost | See the inventory section below; this alone disqualifies some businesses |
| Job estimates vs actuals, job status, job descriptions | Do not convert (jobs themselves become sub-customers or Projects) | Job costing history and in-flight job tracking need a rebuild |
| Memorized transaction groups | Do not convert | Recurring batches must be recreated one at a time |
| Budgets | Only P&L-type budgets come across, with known mapping problems | Re-check every budget line; rebuild balance sheet budgets |
| Progress invoices | Do not convert as progress invoices | Partially billed jobs need manual tracking through the transition |
| Open purchase orders | Convert, but do not reflect items already received; PO-to-bill links must be reconnected manually | Open procurement needs a line-by-line review after migration |
| Commented reports and vehicle mileage history | Do not convert | Export before you migrate if you need them |
| Closed period designations | Do not convert | Re-set the closing date and password in QBO immediately, or prior periods are open to edits |
| Memorized reports | Convert to custom reports on the Advanced tier only | On Simple Start, Essentials or Plus, rebuild every saved report |
For balance: classes and locations convert. Multi-currency preferences and transactions convert on Plus and Advanced. Sales orders, open and closed, convert. Attachments up to 30MB convert. 1099 vendors become Contractors, estimates come across, and jobs arrive as sub-customers or Projects. The tool is genuinely good at lists and plain transactions. It is the structured, workflow-carrying data that gets dropped.
The five failure modes that hurt in practice
1. The overwrite. This is the dangerous one. Re-running a migration completely overwrites the target QBO company, destroying anything entered since the first attempt. There is no selective redo and no partial rollback. The pattern that produces disasters: a first migration looks mostly fine, the team starts entering live transactions, someone notices a problem two weeks in and re-runs the tool to “fix” it, and two weeks of real bookkeeping vanishes. Treat the migration as a one-shot event. If the first pass is wrong, decide deliberately, with everyone informed, whether to redo it before any live work begins.
2. Sales tax lands in the wrong place. Filed-but-unpaid sales tax liabilities move out of A/R and A/P into Suspense accounts, and migrated tax payments can land in the wrong filing periods. The fix is to delete and recreate the affected payments in QBO’s Sales Tax Center, but you have to know to look. If you file in multiple jurisdictions, put sales tax at the top of your validation checklist.
3. The first reconciliation is wrong. Reconciliation history does not transfer, so the opening balance on your first QBO bank reconciliation is usually off. Cleared status does carry on individual transactions, which is what makes the first reconciliation recoverable rather than impossible, but budget real time for it and do not let an untrained user “fix” the opening balance with an adjustment entry.
4. The cash-basis false alarm. Intuit is explicit that accrual-basis reports should match between Desktop and Online after migration, and that cash-basis reports may not, because the two products compute cash basis differently. A large share of the “the migration corrupted my books” posts you will find online are someone comparing cash-basis P&Ls. Validate on accrual. Compare the accrual balance sheet, accrual P&L, A/R aging and A/P aging as of the migration date before you conclude anything is broken.
5. Everything around the file. Integrations do not carry over and must be reconnected or rebuilt against QBO’s API. User permissions do not map; QBO’s roles are structured differently, so access must be redesigned, not copied. Both are project work items in their own right, and both are routinely left off the estimate.
The two hard walls, and the tier trap
Two QBO limits are absolute, and hitting one mid-project is the classic scope blowup.
First, the chart of accounts is capped at 250 on Simple Start, Essentials and Plus. Mature Desktop files, especially ones that have absorbed a decade of subaccounts, blow past 250 routinely. Second, classes and locations are capped at 40 combined on Plus. A Desktop file with 60 classes cannot land on Plus, full stop. Either you consolidate classes before migrating or you pay for Advanced.
The softer walls matter too: custom fields run 1, 4, 4 and 12 across the four tiers, and billable users run 1, 3, 5 and 25. Add the earlier point that memorized reports only convert on Advanced, and a pattern emerges. The tier you priced the switch against is often not the tier your file actually fits. Run the counts before you commit: in Desktop, press F2 or Ctrl+1 and count your accounts, classes and locations against the tier you intend to buy. It takes five minutes and prevents the most common mid-project surprise we see.
While you are in that F2 window, note your target count. Intuit’s tool accepts files up to 5 GB and 4,000,000 targets, but Intuit itself warns that files over 750,000 targets may show data discrepancies, especially in inventory. Other constraints worth knowing before you start: you have a 60-day window from QBO account creation to complete the migration; your Desktop version must be at 2021 R17, 2022 R15, 2023 R12 or 2024 R6 or newer; the file must be local, not in a hosted environment; the target QBO company must not already have payroll enabled; and the migration dashboard keeps your files for only seven days.
One warning belongs in every guide and appears in almost none: do not run Condense Data to get under the size limits. Condense irreversibly destroys transaction detail and replaces it with monthly summaries. If your file is too big, that is a conversation about strategy, not a checkbox to click.
Inventory: the honest section
If you build assemblies in Desktop, read this twice.
Inventory assemblies do not survive migration as assemblies. Intuit’s own phrasing is that they become a single inventory item on sales transactions: components appear as detail lines on invoices, and the assembly build structure is effectively lost. If your business builds products from components, the thing your operation runs on does not exist on the other side of this migration in any usable form. QBO’s native inventory does not do assembly builds the way Desktop does, and no amount of post-migration cleanup recreates them. Businesses in this position either stay on Desktop, move to Enterprise, or pair QBO with a third-party inventory system, which is a new implementation project rather than a migration.
The default setting in the migration tool is “Don’t move Inventory.” That default exists for a reason.
There is also a costing method issue. Desktop Pro and Premier value inventory at average cost; QBO has historically forced FIFO. If your costing method changes, that is a change in accounting method, potentially requiring IRS Form 3115. We are not tax advisors and will not tell you how that applies to you. We will tell you it must be a conversation with your CPA before migration, not a discovery after.
And a note for Mac users: Mac inventory will not import at all. It requires conversion through a Windows Desktop file first, which adds a step and a Windows licence to your project plan.
Stay, go, or go somewhere else
Treat “do not migrate” as a legitimate outcome of this analysis, because it is.
Staying on Desktop is viable for a low-volume, single-entity business. After support ends, bank feeds stop, but you can download statements from your bank, convert them to .QBO, IIF or CSV, and import manually. For many small operations that is 30 to 60 minutes a month. Existing Plus subscribers can also simply keep renewing and retain full support. Staying stops being viable if you need payroll, in-product card payments or multi-currency, all of which deactivate on unsupported versions, or if you have any compliance, insurance or customer requirement around running patched software. That last one is the real long-term risk of staying: an unpatched application holding your financial data. We would not run one indefinitely, and we do not recommend our clients do either.
Going to QBO makes sense when the business genuinely needs what Online is for: multiple people in the books from multiple places, an accountant with live access, bank feeds that just work, and a modern integration ecosystem. If your file is a service business without assemblies, without complex jobs and without multiple A/R accounts, the migration is well-trodden and the losses in the table above may cost you nothing you actually use.
The alternatives deserve a look before you commit. Desktop to Enterprise is a much cleaner data path because it is the same data model; assemblies, price levels and jobs come along intact, and Enterprise is still sold to new customers. For businesses whose problem is Pro or Premier’s limits rather than the desktop form factor, it is often the better move. Xero is the other credible destination: it offers free conversion via Jet Convert, but only for the current and prior fiscal year, and payroll explicitly does not migrate. Shorter history plus a payroll rebuild is a real price; for some businesses it is worth paying, but go in knowing it.
Two one-way doors to respect regardless of direction. Multi-currency, once enabled in QBO, cannot be disabled, so do not switch it on speculatively. And if you run multiple entities, understand that there is no batch path: one Desktop file means one QBO subscription and one separate migration, each with its own 60-day window and its own monthly fee. A holding company with four entities is signing up for four migrations and four subscriptions. This blindsides more multi-entity owners than any other line in this article.
What a clean migration project actually looks like
The file transfer itself takes minutes to about 24 hours. The project around it is where the real hours live, and honest numbers look like this: a simple service business with no inventory and no payroll history needs 8 to 15 hours of work spread over one to two weeks. A typical small business with payroll and some inventory needs 25 to 50 hours over three to six weeks. Re-entering year-to-date payroll for around 50 employees is several days of careful work by itself.
- Assess first. Run the file against the table above, check F2 for targets and list counts, confirm the tier fits the 250-account and 40-class walls, and inventory every integration and every user’s actual permissions. Decide stay, go or alternative on evidence.
- Clean up in Desktop. Merge duplicate accounts, retire dead list entries, resolve old open transactions. Whatever you migrate, you keep. Never use Condense to do this.
- Time it around payroll. Best practice is to migrate two to three days after the final Desktop payroll run, so the current-year lump sums are as complete as possible and the year-to-date re-entry is a single clean exercise.
- Validate on accrual, on day one. Compare the accrual balance sheet, accrual P&L and both agings against Desktop as of the migration date, before anyone enters a live transaction. Check sales tax balances against Suspense specifically. If it is wrong, this is the only cheap moment to re-run.
- Run 30 days parallel. Keep Desktop as the reference while the team works in QBO. Reconcile the first month in both. Only then stop entering in Desktop, and keep the file forever as your archive of the audit trail, payroll detail and reconciliation history that did not come across.
- Rebuild the perimeter deliberately. Recurring transactions, report packages, user roles, integrations and the closing date all need to be recreated by hand. Put them on the plan with names and dates, or they will surface as small emergencies for a quarter.
Notice what this list is: an accounting-controls project with a software step in the middle. The businesses that get hurt are the ones that treat it as a software step alone.
Questions we get asked
What does not migrate from QuickBooks Desktop to Online?
Payroll detail beyond the current year, QuickBooks Time data, the audit trail, past reconciliation reports, inventory assembly structures, price levels, billing rate levels, sales reps, units of measure, manufacturer’s part numbers, job estimates versus actuals, memorized transaction groups, progress invoice status, vehicle mileage, closed period designations, and all but one A/R and one A/P account. Integrations and user permissions must be rebuilt from scratch. The table above has the practical consequence of each.
Is QuickBooks Desktop being discontinued?
No full sunset has been announced. Intuit stopped selling Pro Plus, Premier Plus and Mac Plus to new US subscribers on September 30, 2024, but existing subscribers can renew indefinitely, and Enterprise is still sold to new customers. Support ends per version on a rolling three-year schedule: the 2023 versions ended May 31, 2026, and 2024 and Enterprise 24.0 run through September 30, 2027.
Can I keep using Desktop after support ends?
Yes. The program keeps opening and working; bookkeeping, invoicing, reporting and printing continue indefinitely. What stops is the connected layer: bank feeds, payroll, payments, live support and security updates. The security updates are the part to take seriously if you plan to stay for years.
How long does the migration take?
The transfer runs minutes to about 24 hours. The project is 8 to 15 hours over one to two weeks for a simple service business, and 25 to 50 hours over three to six weeks for a typical small business with payroll and inventory, plus roughly 30 days of parallel running before you retire Desktop.