The 7-Day Average Stay Test for Short-Term Rentals

How the average is calculated

Total nights rented during the year, divided by the number of separate rental periods. If you rented 180 nights across 40 bookings, the average is 4.5 days and you are clear. If those same 180 nights came from 20 bookings, the average is 9 days and you are not.

It is the actual average across the year, not your advertised minimum and not a typical stay. Vacant nights are not counted as rental days — the divisor is bookings, not calendar time.

What quietly pushes a property over

The properties that fail rarely look like long-term rentals. They look like this:

• A month-long winter booking taken to fill a slow season, against otherwise short stays.
• A relocation or insurance guest placed for six weeks.
• A slow year where a handful of long bookings make up most of the nights and short stays never materialised.
• Switching platforms or raising the minimum stay mid-year without recalculating.

Check it during the year, not after

The test is annual, but you find out whether you passed only once the year is closed — by which point nothing can be changed. Owners who watch the running average through the autumn still have the option to decline or restructure a long booking. Owners who check in April do not.

There is a related point worth knowing: a second exception exists where the average is 30 days or less and significant personal services are provided. It is narrower and more fact-dependent, and it is firmly a conversation to have with your CPA rather than something to assume.

Keep the booking record too

Your platform’s booking history is the evidence for this test, so keep an export of it with your hours log. A tax position that depends on two conditions needs documentation for both, and the average-stay calculation should be reproducible from your own records rather than from a login you might lose.