It is almost never the arithmetic
When a return comes back, the instinct is to check the calculation. That is usually the one part that was fine. Tax per jurisdiction is a formula, and a formula applied to bad inputs produces a confident, well-formatted wrong answer.
The inputs are the problem, and there are only two of them: how many miles you ran in each jurisdiction, and how much fuel you bought in each jurisdiction. Everything else is derived. If either input is incomplete, the return is a guess that happens to be laid out neatly.
Distance records that actually support a number
An ELD records miles. That is not the same as producing a distance record that supports a filing. What a jurisdiction wants to see is that total distance reconciles to individual trips, and that the jurisdiction split is derived from where the truck actually was rather than estimated after the fact.
Common failures:
- Gaps. The odometer at the end of one trip does not match the start of the next, and nothing explains the difference. Personal conveyance, yard moves and deadhead all still happened on real roads in real states.
- Estimated splits. Miles apportioned to jurisdictions by assumption rather than by route. It is usually close. Close is not evidence.
- Missing non-revenue miles. Empty miles are taxable miles. A return that only counts loaded miles understates distance everywhere.
- Unit mismatch. Distance filed against a vehicle that was not the vehicle that ran it, usually after a truck swap mid-quarter.
Fuel receipts that count
A fuel receipt earns you tax-paid credit in the jurisdiction where the fuel was bought. To do that it generally has to show the date, the seller and location, the number of gallons, the fuel type, and the vehicle it went into.
A card statement line that reads TRUCK STOP 4471 — $612.40 shows none of that. It proves you spent money. It does not prove you bought 148 gallons of diesel in Ohio and put them in unit 12.
If you take one operational change away from this: keep the itemised receipt, not just the card record. Fuel-card reports are usually sufficient when they carry gallons and location per transaction. Check yours actually does, before the quarter you need it.
What the auditor is really testing
An IFTA auditor is not hunting for a number they disagree with. They are testing a chain: can the figure on the return be traced back to a source record, for every jurisdiction, for the whole period?
They will typically sample. A handful of trips, a handful of receipts, and a check that the sample reconciles to the total. If the sample holds, the return holds. If the sample does not, the problem is not that one trip — it is that your records cannot be relied on, which puts the entire filing in question.
That is the part carriers find hardest to accept. A small, honest gap in evidence can invalidate a large and substantially correct return, because the auditor is assessing the system, not the individual number.
The fix is boring
Reconcile every quarter while the quarter is fresh. Match total ELD distance to the sum of your trips, and match fuel gallons to itemised receipts. Any difference you cannot explain in the same week it appears is a difference you will not be able to explain in three years.
Keep the records for the full retention period your base jurisdiction requires, and keep them in a form somebody other than you can follow. An audit that arrives after a driver has left or a phone has been replaced is an audit against whatever survived.
None of this requires software or expertise. It requires it to be somebody's job, every quarter, without exception — which is exactly the thing that does not happen when the owner is also the driver.