Eduints
Audit & Assurance

Revenue Cut-Off Testing in an Audit

An invoice from the last day of the year: 31 March, ₹48 lakh. The gate register shows the goods left on 3 April. Would you pass this sale into this year’s revenue?

Written by the Eduints teamPublished 1 October 2026

The invoice date is not the test

No — the goods left in April. Revenue from selling goods is recognised when the significant risks and rewards of ownership pass to the buyer. Until goods leave, the company still holds them; the invoice date is just a date someone typed. And the customer having already paid doesn’t change the answer either — payment and ownership are different things. A customer paying in advance creates a liability, not revenue.

Test both sides of the year end

A cut-off test has two sides. Before the year end: take sales recorded in the last few days of March, and check whether each one really left by the 31st — this finds sales recorded too early. After the year end: take sales recorded in the first days of April, and check whether any really left in March — this finds sales recorded too late. Testing only one side is testing only half the risk.

The correction has two halves

A full test here finds ₹1.2 crore of sales recorded in March for goods that left in April. The correction: reverse the sale and the receivable, and, because the goods were still in the warehouse on the 31st, put the stock back at cost (₹0.9 crore) and reverse the cost of sales. Reversing revenue while forgetting the inventory is one of the most common slips here — it overcorrects profit, since it removes the sale without removing its cost.

Net effect: revenue overstated by ₹1.2 crore, cost of ₹0.9 crore due back into inventory, so profit before tax was overstated by ₹0.3 crore — a large error, well above performance materiality on its own.

What actually decides when risk passes

Three signs: the goods have left — dispatch is recorded and the carrier has them. The contract terms say when risk passes — on dispatch or on delivery, and that decides the date, not the invoice. And the price is fixed with collection reasonably certain — if goods can be returned for any reason, or the price isn’t agreed, the sale isn’t complete whatever the invoice says.

Watch for bill-and-hold arrangements — goods invoiced but kept in the seller’s warehouse at the customer’s request. These are rarely acceptable revenue recognition and only under strict conditions; treat one as a red flag and ask for the conditions in writing. Goods sent on a sale-or-return basis aren’t a sale yet either — if the customer can return them at will, the risk hasn’t passed.

Credit notes after year end are evidence

If credit notes are issued in April for goods sold in March, that’s evidence the March sale was, at least partly, not what it seemed — information about conditions that existed at year end, which means the revenue should be reduced in March, not treated as a clean April transaction.

Designing the test

For the last five days of March, take all sales above the clearly trivial threshold and check dispatch was on or before the 31st. For the first five days of April, the same, checking dispatch was after. Review April’s credit notes relating to March sales. The windows are short because errors cluster near the year end, and selection is every item above the threshold — not a sample — because there aren’t many, and a sample could miss the few that matter.

When management pushes back

“The customer had agreed to take the goods — it’s a March sale in substance.” The response is two factual questions: what do the terms say about when risk passes, and where were the goods on 31 March? A commitment is an order. It becomes a sale only when the goods and the risk actually move — and those questions have answers on paper, not opinion.

If the client offers to correct just the one error found and call it a mistake, the right response is to continue testing the whole population anyway — one error found means the process that produced it may have produced others. The test stops when the planned population has been covered, not when the client fixes what’s already been caught.

Mistakes that undermine a cut-off test

  • —Testing cut-off only for sales dated before the year end, missing the April-side errors.
  • —Reversing revenue but forgetting to reinstate inventory and cost.
  • —Relying on the invoice date as the date of sale, instead of dispatch and the contract terms.
  • —Ignoring credit notes issued after year end that relate to year-end sales.

This guide illustrates standard revenue cut-off testing using a fictional teaching case. It is practical educational content, not professional audit or accounting guidance — a real engagement applies the specific revenue-recognition standard in force for that entity.

Go deeper with the full engagement

This cut-off testing framework is Lesson 19 of Advanced Audit & Assurance, one module inside a full fictional engagement — from accepting the client through planning, testing, and forming the final opinion.

Have a question this guide didn’t answer? See the full FAQ or contact us directly.