The short-term rental exception is a rule in the passive activity loss regulations: a property whose average guest stay is seven days or less is not a “rental activity,” so if the owner materially participates, its losses are non-passive and can offset wages and other ordinary income. Online it goes by “the short-term rental loophole” or “the STR loophole.” It is not a loophole. It is a definition in a regulation, and it comes with a participation test that decides whether it works.
Every year around October, the same conversation happens in our office. A physician in San Antonio, an engineer in Austin, a sales executive in Dallas: someone with a large W-2 and a large tax bill has read that buying a short-term rental before December 31 can wipe out a meaningful slice of that bill. Sometimes they’ve read it on a forum. Sometimes a friend did it. Almost always, the version they heard skipped the part that decides whether it works.
The strategy is real. It has been in the regulations since 1988, and the IRS has issued guidance on it. But it is not a “buy a cabin, deduct everything” trick, and the participation test is where most of the failed versions fall apart. This article walks through the actual mechanics: what the regulation says, what “material participation” means in hours and documentation, why cost segregation is the engine that makes the numbers large, and the handful of ways the whole thing quietly unwinds.
Why rental losses are usually trapped
Start with the default rule, because the exception only makes sense against it. Under Section 469, rental real estate is passive by definition, regardless of how many hours you put in. Passive losses can only offset passive income. If your rental shows a $60,000 tax loss and your only other income is a salary, that loss doesn’t reduce your salary. It is suspended, carried forward, and released either against future passive income or when you sell the property.
There is a small-landlord allowance: up to $25,000 of rental losses can offset other income if you “actively participate,” which is a low bar. But that allowance phases out between $100,000 and $150,000 of modified adjusted gross income and has never been indexed for inflation. For the households we typically work with, it is fully gone.
That leaves two escape hatches. The first is Real Estate Professional Status, which requires more than 750 hours per year in real property trades or businesses and more than half of all your working hours in them. For a full-time physician or executive, that test is essentially impossible, and a spouse who qualifies has to clear it on their own hours. The second escape hatch is the subject of this article.
What the regulation actually says
The passive loss regulations define “rental activity” more narrowly than everyday English does. Under Treasury Regulation 1.469-1T(e)(3), an activity is not a rental activity if any of six exceptions apply. The two that matter for short-term rentals:
The seven-day exception. If the average period of customer use is seven days or less, the activity is not a rental activity. Full stop. No services requirement, no professional-status test. A property with an average booking length of 3.8 nights is, for passive-loss purposes, a trade or business, not a rental.
The thirty-day exception. If the average period of customer use is 30 days or less and the owner provides significant personal services (daily housekeeping, concierge, meals, the kind of thing a hotel does), the activity is likewise not a rental activity. This one is narrower in practice because “significant personal services” excludes the routine stuff: cleaning between guests, repairs, utilities, and checking people in don’t count.
Once an activity falls outside the rental definition, it is treated like any other trade or business under Section 469. Which means the question is no longer “are you a real estate professional?” It is “do you materially participate?” That is a far easier standard to meet, and it is the entire reason the strategy exists.
Two clarifications the forums usually miss. First, the seven-day figure is an average, computed by dividing total days rented by the number of separate rental periods for the year. A property with mostly weekend stays and one 21-night winter booking can still clear it; a property with mostly monthly tenants and a few weekends will not. Second, the test is applied property by property, per year. A property that averages six nights in 2026 and nine nights in 2027 is non-rental in the first year and rental in the second, with the losses treated accordingly.
Material participation: the part that decides everything
Material participation is defined in Regulation 1.469-5T through seven alternative tests. You only need to satisfy one. For short-term rentals, three of them do almost all the work:
Test 1: more than 500 hours. You participate in the activity for more than 500 hours during the year. For a single property, this is unusual and hard to document credibly, but it is the cleanest path for an owner running several units.
Test 2: substantially all of the participation. Your participation constitutes substantially all of the participation of all individuals in the activity, including non-owners. This is the test that a property manager destroys. If a management company handles listings, pricing, guest communication, and turnover, their hours dwarf yours, and you fail.
Test 3: more than 100 hours and not less than anyone else. You participate for more than 100 hours, and no other individual (including cleaners, handymen, and co-hosts) participates more than you do. The comparison is person by person, not against a company’s combined staff time. This is the test most W-2 owners actually use. A hundred hours across a year is roughly two hours a week: listing management, dynamic pricing, guest messaging, restocking, scheduling and inspecting turnovers, handling the inevitable 11 p.m. lockout. It is achievable. But note the second half of the test: if your cleaner logs 140 hours across 70 turnovers, your 110 hours do not clear it.
The remaining four tests (significant participation activities aggregating over 500 hours, five of the prior ten years, personal service activities, and a facts-and-circumstances test that itself requires more than 100 hours) rarely fit a new short-term rental owner.
Three points on how hours are counted, because they matter more than the tests themselves.
Spousal hours combine. Under Section 469(h)(5), participation by your spouse counts as your participation for material participation purposes, whether or not you file jointly and regardless of who holds title. In a community property state like Texas, this fits naturally: if one spouse works a demanding W-2 job and the other runs the property, the household clears the test together. (This is one of the sharpest contrasts with REPS, where the 750-hour and 50% tests must be met by one spouse alone.)
Investor-type hours don’t count. Reviewing financial statements, monitoring operations in a non-managerial capacity, and preparing your own summaries for tax purposes are excluded unless you are directly involved in day-to-day management. Travel time to the property is contested and frequently disallowed.
Contemporaneous records win; reconstructions lose. The regulation permits “any reasonable means” of substantiation, but the case law is consistent: a log built after the fact from memory and calendar guesses is routinely rejected, and “ballpark guestimate” has become the Tax Court’s shorthand for the losing taxpayer. Lucero v. Commissioner (2020) is close to a case study for this article: a short-term rental with an average stay under seven days, a property management company running day-to-day operations, and an owner whose hours log was reconstructed from receipts only after the IRS had already challenged the return. The Tax Court threw out his drive time between Sacramento and the property as personal commuting, discounted the rest of the log because some entries were padded (two hours to buy coffee filters), and disallowed the losses. A simple running log, dated entries, what was done, how long, kept as you go, is the single most valuable document in this strategy.
The log has to cover other people’s hours too. Tests 2 and 3 are comparative, and the burden of proving the comparison is yours. In Pohoski v. Commissioner (1998), the taxpayers cleared the test on one of their two Hawaiian condos but failed it on the other, because they offered no evidence of how many hours the resort’s front desk and cleaning staff had spent on that property. Your cleaner’s invoices, your handyman’s time, and any co-host’s activity belong in the same file as your own hours.
Why cost segregation is the engine
Clearing the passive rules makes a loss usable. It does not make a loss large. On a standard depreciation schedule, a $600,000 building depreciates over 27.5 years, producing roughly $21,800 per year. Against $50,000 of rental income, that is not a loss at all.
Cost segregation changes the size of the number. A study reclassifies components of the building (flooring, cabinetry, appliances, dedicated electrical, land improvements like driveways and fencing) from 27.5-year property into 5-, 7-, and 15-year classes. Under the OBBBA bonus depreciation reset, property in those shorter classes that is acquired and placed in service after January 19, 2025 is eligible for 100% bonus depreciation, meaning the entire reclassified amount is deducted in year one. We covered the mechanics and the break-even math in our cost segregation article and the OBBBA changes in the bonus depreciation reset piece; the short version is that for a furnished short-term rental, studies commonly reclassify a meaningful share of the building cost because these properties are furniture-, fixture-, and site-improvement-heavy.
One point of genuine disagreement among practitioners deserves a flag before the example. Residential rental property depreciates over 27.5 years; nonresidential property over 39. The definition of residential rental property excludes units used on a transient basis, and a number of cost segregation firms and CPAs take the position that a short-term rental is therefore nonresidential and belongs on a 39-year schedule (which also opens the door to qualified improvement property treatment for later renovations). Others, reading the statutory language as aimed at hotels and multi-unit establishments, keep a single-family short-term rental at 27.5. The choice changes the annual depreciation on the building shell and how the cost segregation study is built, and it needs to be settled between the study provider and the tax preparer before the return is filed, not after. The example below uses 27.5 years; on a 39-year schedule the year-one loss would be modestly smaller.
Here is the combination, in a stylized example. A married couple in Austin with $480,000 of combined W-2 income buys a furnished Hill Country short-term rental in September for $750,000. Their allocation puts $150,000 in land (never depreciable) and $600,000 in the building and its components. A cost segregation study reclassifies $150,000 into 5-, 7-, and 15-year property. With 100% bonus depreciation, the entire $150,000 is deducted in the year of purchase, on top of regular first-year depreciation on the remaining $450,000 of building.
The property generates $18,000 of net rental income before depreciation in its first partial season. The year-one tax loss comes to roughly $137,000. Because the average stay is under seven days and one spouse logged 130 documented hours (more than the cleaner’s 90), the activity is non-rental and materially participated in. The $137,000 loss is non-passive and offsets W-2 income. After the $32,200 standard deduction, the couple’s taxable income of roughly $448,000 sits in the 32% bracket, which for 2026 joint filers runs from $403,551 to $512,450. The loss clears about $44,000 out of the 32% bracket and the remaining $93,000 out of the 24% bracket, for roughly $36,000 of federal income tax not paid this year, with no Texas income tax to consider.
That number is why the strategy gets so much attention. Three things about it deserve equal attention.
Three gates, and the loss must clear all of them
The seven-day rule and material participation are gate one. Two more sit behind it, and each can catch a loss that has already cleared the passive rules.
Gate two: basis and at-risk. You can only deduct losses up to your basis in the activity, which for a directly owned property is generally your cash invested plus debt you are personally liable for. Qualified nonrecourse financing on real estate generally counts as at-risk. In practice this gate rarely bites on a conventionally financed purchase, but it can on seller-financed deals and inside entities.
Gate three: the excess business loss cap. Even a fully non-passive loss faces a final ceiling. Under Section 461(l), which OBBBA made permanent and re-based, aggregate business losses that can offset non-business income (wages, interest, capital gains) are capped at $256,000 for single filers and $512,000 for joint filers in 2026, indexed thereafter. Anything above the cap becomes a net operating loss carried to the next year. A $137,000 loss is well inside it. A couple buying two large properties in one December, or an owner stacking a short-term rental loss on top of a large operating-business loss, can hit it.
Where it quietly unwinds
The mechanics above are the version that works. These are the versions we see not work, roughly in order of frequency.
The property manager. Full-service management is the most common reason the strategy fails. If a manager handles the listing, pricing, and guest communication, the owner cannot meet test 2 (substantially all). Test 3 is a closer call because it compares your hours to each other individual, not to the management company as a whole; a firm whose work is spread across several people, none of whom exceeds your hours, does not automatically break it. In practice, though, a dedicated account manager usually out-hours a W-2 owner, and proving otherwise requires records of the firm’s staffing you rarely have. Self-managing through a platform, with hired cleaners and a handyman, is compatible with the tests. Handing the whole thing to a management company generally is not.
Bought in December, listed in January. Depreciation, including bonus, starts when the property is placed in service: furnished, listed, and actually available for booking. A closing on December 28 with the first listing going live in the new year puts the entire deduction in the following tax year. The purchase date is not the test.
Bought in December, and the hours aren’t there. The material participation tests are annual and are not prorated for a short year. In Gregg v. United States (2000), a federal district court refused to annualize the 500-hour test for an activity that began in November, reasoning that proration would invite exactly the year-end purchases the passive loss rules were written to stop. A late-year buyer is realistically working with test 3: more than 100 documented hours before December 31, and more than anyone else. Furnishing, listing, pricing, and hosting the first guests can get there, but it has to be logged, and it has to happen before year-end.
Personal use. If your own use of the property exceeds the greater of 14 days or 10% of the days it is rented at fair value, Section 280A treats it as a residence and limits deductions to rental income. No loss at all. Families who buy a lake house “as a short-term rental” and then spend summer weekends there need to run the day count before modeling a six-figure loss.
The average creeps up. A property that shifts toward monthly stays, common when an owner tries to reduce turnover work, can cross the seven-day average and become a rental activity again. The material participation you documented no longer matters; you are back to REPS or passive. The earlier year’s treatment stands on its own facts and does not need to be amended. If the shift in use also changes the property’s recovery period (the 27.5 versus 39-year question discussed above), depreciation adjusts prospectively under the change-in-use regulations; a Form 3115 is for the different situation where the wrong recovery period was used from the start.
The hours were reconstructed. See above. If the log doesn’t exist, the strategy doesn’t exist.
The next tax return. The loss is a year-one phenomenon. In year two the property is likely profitable on paper (small remaining depreciation, a full season of income). Both the seven-day test and material participation are determined fresh each year, and the character of that year’s income or loss follows that year’s facts. Keep participating, and the income is non-passive. Stop participating while the average stay remains under seven days, and the activity is still non-rental but its income is passive for that year, which means it can absorb suspended passive losses from other activities. (One trap: if you still put in more than 100 hours that year without clearing a material participation test, a recharacterization rule treats that income as non-passive anyway, and the offset disappears.) Neither outcome disturbs the year-one loss, which was decided on year-one facts.
Two things this strategy is not
It is not a deduction, it is a deferral. We said this about cost segregation and it applies here with equal force: bonus depreciation pulls deductions forward, it does not create them. When the property sells, the reclassified 5-, 7-, and 15-year property is recaptured as ordinary income under Sections 1245 and 1250 (personal property and land improvements respectively), and the building depreciation is recaptured at up to 25% under Section 1250. The $36,000 in the example is tax not paid this year. Whether it is a good trade depends on the hold period, the bracket at sale, and whether the exit is a taxable sale, a 1031 exchange, or a step-up at death. We lean toward modeling the full arc before purchase, not just year one.
It is not automatically self-employment income. A frequent objection is that a non-rental activity must go on Schedule C and owe 15.3% self-employment tax, which would eat most of the benefit. The IRS addressed this in Chief Counsel Advice 202151005: the seven-day average by itself does not create self-employment income. Only substantial services to guests (the hotel-style services from the 30-day exception) push the activity onto Schedule C and into SE tax. A short-term rental with turnover cleaning, linens, and a stocked coffee bar stays on Schedule E. This is worth getting right, because it also determines whether the eventual income is subject to the 3.8% net investment income tax: non-passive income from an activity you materially participate in is generally outside NIIT, which begins at $200,000 of MAGI for single filers and $250,000 for joint filers. And because the activity is a trade or business rather than a rental, its income in profitable years may qualify for the Section 199A deduction; our QBI article covers the thresholds and the wage-and-property limits that apply above them.
How we think about it
One thing worth saying plainly, because the marketing around this strategy rarely does: it does not create wealth. The tax deferred in year one is always smaller than the price of the property that generated it, and a property bought for the deduction rather than for its returns, location, and fit with the family’s life is a bad purchase with a good tax return attached. The question comes in that order: is this a property you would want to own and operate anyway? Only then does the tax treatment matter.
The short-term rental exception is a legitimate, well-documented piece of the passive loss rules, and for the right household it can move real dollars. But it is a household with a specific shape: a large W-2 or professional income, a spouse or the owner themselves with roughly two hours a week to actually run the property, no interest in a full-service manager, a property large enough for a cost segregation study to matter, a hold period long enough for the deferral to be worth the recapture, and personal use that stays well inside the 14-day line.
When those things line up, the strategy works as advertised. When one of them doesn’t, the loss lands as passive, gets suspended, and the owner has bought a vacation property with a management job attached and no tax benefit until they sell it. The difference is knowable in advance, which is why we model it: the hours, the average stay, the personal-use days, the year-one loss against the three gates, and the recapture at exit, before anyone goes under contract.
If you already own a property and are wondering whether it qualifies, or you converted a primary residence and are weighing long-term versus short-term use, our article on turning a home into a rental covers the basis and Section 121 questions that come first.
Common Questions
Does the seven-day rule mean every Airbnb-style property is non-passive?
No. Clearing the seven-day average only removes the property from the “rental activity” definition. The loss is still passive unless you also materially participate, which for most owners means more than 100 hours in the year and more than any other individual, including cleaners and co-hosts, with contemporaneous records to prove it.
Can my spouse’s hours count if the property is only in my name?
Yes. Under Section 469(h)(5), a spouse’s participation is attributed to you for material participation purposes regardless of title or filing status. Texas community property rules don’t change the answer, but they do make the household framing natural.
Do I owe self-employment tax on a short-term rental?
Generally not. The IRS’s position in Chief Counsel Advice 202151005 is that a short average stay by itself does not create self-employment income; only substantial hotel-style services to guests do. Routine cleaning between stays, linens, and maintenance keep the activity on Schedule E.
What if I use a property manager but still do a lot of the work?
It is difficult. Test 2 requires substantially all participation, which a full-service manager rules out. Test 3 requires more hours than any other single individual, so a manager whose work is spread across several people is not automatically fatal, but the person assigned to your property will usually out-hour you, and you would need their records to show otherwise. Owners who want the exception typically self-manage the listing and guest side and hire only task-specific help.
What happens to the loss if I fail the tests?
It becomes a passive loss, suspended and carried forward. It is not lost: it offsets future passive income and is released in full when you dispose of the property in a fully taxable transaction. Whether that timing is acceptable is a question to answer before buying, not after.
How much can I use the property myself?
If your personal use exceeds the greater of 14 days or 10% of the days the property is rented at fair value, Section 280A treats it as a residence and limits deductions to rental income, which eliminates the loss. Days spent at the property primarily on repairs and maintenance generally do not count as personal use, but days used by family members or anyone paying less than fair rent do.
Does the excess business loss cap apply to me?
Only if your total non-passive business losses for the year exceed $256,000 (single) or $512,000 (joint) in 2026. A single short-term rental rarely gets there. Multiple properties in one year, or a rental loss stacked on an operating-business loss, can.