Cost segregation risks: what can go wrong and what to check before you pay

By CostSegregationMatch · Tax sources checked October 7, 2026

Short answer: Cost segregation is legal. The IRS even publishes the guide its examiners use to review these studies. The risk is not that it's a trick. The risk is that it doesn't pay off for you, or that a weak study doesn't hold up.

Five things can cost owners real money:

1. You can't use the loss yet. Rental losses are generally "passive." If you lack passive income and do not qualify for an exception, the extra loss can wait instead of reducing this year's tax.
2. Tax can come back when you sell, and the rate can be higher. Gain on some short-life items is taxed at your regular federal rate, up to 37%, limited by the depreciation and gain on those items. The 25% number people quote only covers part of it.
3. A big sale can put you in a higher bracket than the one you deducted in. That can turn an early tax saving into a loss overall.
4. A study with thin support can be cut back in an audit. Then you owe the tax, interest, and maybe a 20% penalty.
5. Your state may not follow the federal rule. California, for one, does not allow federal bonus depreciation.

Here is what that looks like in dollars. In our simplified example of a $1,000,000 rental, with the extra deduction fully usable at 32%, the study saves about $61,700 of federal tax in year one. Sell after five years with the short-life items valued at original cost, and the model shows about $54,000 of extra sale tax. After $3,000 of fees, that is about $4,600 behind in plain dollars but about $5,600 ahead when future amounts are discounted at 5% a year. A higher sale-year rate can turn the result negative. This compares with a hypothetical all-building benchmark. The example and calculator show the assumptions, including the same overall sale price in both scenarios.

Jump to: Risks at a glance · Can I use the deduction? · What comes back when I sell · Calculator · 1031 exchanges · Audit risk · Worksheet · Got an IRS letter?

Already cleared the tax side with your tax preparer? Compare study options.

Cost segregation risks at a glance

Not every risk applies to every owner. Find your rows.

Can you use the deduction this year?

Maybe not. A study can find real depreciation that does nothing for your tax bill this year. Check this before you buy a study, especially if your main income comes from a day job.

Rental losses are generally passive under the tax rules. Passive losses can only offset passive income, with a few exceptions. Here is how the main ones work, from IRS Publication 925:

The $25,000 is a total annual allowance, not an amount for each property. Active participation generally requires at least 10% ownership, counting your spouse's interest, throughout the year and meaningful management decisions. Married people filing separately who lived apart all year have a maximum $12,500 allowance, phased out between $50,000 and $75,000 MAGI. Those who lived together at any time during the year do not qualify. See the Form 8582 instructions.

A few plain-word notes on that table:

• MAGI means modified adjusted gross income. Your tax preparer can give you the number.
• Real estate professional has two time tests. More than half of your personal services in trades or businesses must be in qualifying real estate businesses in which you materially participate, and those real estate services must exceed 750 hours a year. Another full-time job can make the more-than-half test difficult. On a joint return, one spouse must meet both tests separately. Real estate work as an employee generally counts only if you own more than 5% of the employer.
• Material participation means meeting one of the IRS participation tests. More than 500 hours during the year is one test. Your spouse's participation can count toward material participation, even though you cannot combine spouses' hours to meet the real estate professional time tests.
• Running an Airbnb does not settle it by itself. Check average customer use and the applicable participation test. Short stays do not override other loss limits.

A loss that waits is not gone. It can offset passive income later. Suspended passive losses are generally released when you dispose of your entire interest in the activity to an unrelated buyer in a fully taxable transaction. If properties are grouped as one activity, selling one property may not be enough. Basis and at-risk limits are separate checks.

There is also a limit on very large business losses. For tax years beginning in 2026, the excess business loss threshold is $256,000, or $512,000 for a joint return (IRS Revenue Procedure 2025-32, section 3.31). After other applicable loss limits, a noncorporate taxpayer's otherwise allowable aggregate net business loss above that threshold generally becomes a net operating loss carryover for future years, subject to those years' rules. This can apply even to real estate professionals. The Form 461 instructions explain how the limit works.

Four numbers a sales estimate can blur together

A study estimate may show one large number. Ask which one it is.

1. Amount moved to a shorter life. Say $200,000.
2. Extra depreciation this year. How much more you can deduct than without a study.
3. Extra depreciation you can use this year. After the limits above.
4. Tax saved. The difference between your tax with and without the usable deduction. Multiplying by one assumed rate is a simplified estimate.

Example: if all $100,000 of extra usable deduction offsets income that would otherwise be taxed at 32%, the simplified federal income tax saving is $32,000, before other tax interactions. If none of it is usable this year, it saves $0 this year. A deduction that crosses brackets needs more than one rate.

A choice before you claim bonus depreciation

You can do a study and still elect out of bonus depreciation for a tax class. The election generally covers all property in that class placed in service by that taxpayer during the year. It is generally made on the timely filed return, including extensions, for that year. The affected assets then follow the regular depreciation rules. This changes both the early deductions and potential recapture. IRS Publication 946 explains the election. Ask your tax preparer to compare the options; late elections and later changes need the correct procedure.

What happens to depreciation recapture when you sell?

When you sell at a gain, some of that gain can be taxed under depreciation recapture rules. A study does not create all the tax on a sale. It changes the deductions taken before the sale and how the later gain is taxed. No gain on an item generally means no depreciation recapture on that item in a normal sale.

Why "recapture is 25%" is only half right

A study sorts your building into buckets. Each bucket is taxed its own way at sale.

The recovery period alone does not decide the recapture category. Have your preparer confirm which assets are section 1245 and which are section 1250. Those classifications depend on the assets, not on whether you bought a study. These are federal individual rules for property held more than one year; other gains, losses, and the section 1231 lookback can change the final tax treatment. The 25% rate is a maximum, not a flat tax on every dollar of building gain.

Sources: IRS Publication 544, Form 4797 instructions, IRS Topic 409, and the IRS 2026 rates.

Recapture is limited to your gain on each item

This is the part a flat recapture estimate can miss.

Say a study put $60,000 into carpet and appliances, and you wrote off all $60,000. Five years later you sell. A fair share of the sale price for that worn carpet and those used appliances is $15,000.

• Your tax basis in those items: $0
• Your gain on those items: $15,000
• Recapture at your regular rate: $15,000, not $60,000

The other $45,000 is not ordinary recapture on those items. It does not automatically become taxable gain somewhere else. If the whole property sells for the same total price, assigning less of that price to these items assigns more to the remaining property. That can create more gain there, with the tax treatment depending on its basis, depreciation, and the other sale rules. The example below holds the total sale price fixed.

Two cautions. The split of the sale price has to match what the items are really worth, and you need records to back it up. Used items are not worth zero just because you wrote them off. This is a job for your tax preparer.

Skipping depreciation generally does not preserve your tax basis: depreciation you were allowed to claim can still reduce it. For ordinary section 1245 recapture, proof of a lower amount actually allowed can affect the calculation. Have your preparer reconcile missed deductions and the depreciation record before calculating sale tax.

Worked example: a $1,000,000 rental

Here are the numbers we used. They are made up for the example. They are not typical results.

• Price $1,000,000. Land $200,000. Building $800,000. Residential rental.
• The study moves $160,000 to 5-year section 1245 items and $40,000 to 15-year land improvements classified as section 1250 property.
• Assume the eligible components were acquired and placed in service after January 19, 2025, with no exclusion or election changing the 100% bonus treatment (IRS, January 14, 2026).
• A flat 32% rate on every dollar of the modeled deductions and ordinary gain, and a 15% rate on other modeled long-term gain. The extra deduction is fully usable each year.
• $3,000 for the study and extra tax work. That is our assumption. Real quotes vary. The model subtracts it as a cash cost and does not model a tax deduction for the fee.
• The same whole property sells for $1,000,000 in both scenarios, with $200,000 still allocated to land and no selling costs. The remaining price is allocated to the other assets at their assumed fair values. The benchmark keeps the full $800,000 in building basis. This is a hypothetical all-building comparison, not a rule that all assets become building property if you skip a study. There is enough building gain to apply the model's depreciation-related gain categories.
• Federal tax only. This is a full-year timing approximation, not a return calculation: it omits building mid-month timing and sale-year partial-year adjustments. Fees are paid at the start, tax effects occur at each year-end, and the present-value calculation discounts them at 5% a year. State tax, net investment income tax, bracket changes within a year, other gains and losses, and special elections are not modeled. The timing omissions can materially change short-hold or nonbonus results.

"Extra tax when you sell" means the extra tax compared with the hypothetical all-building benchmark at the same overall sale price. The study does not itself determine an asset's recapture category; your preparer must establish the correct classifications, basis, and depreciation in both actual scenarios. Rows may be off by $1 from rounding.

Two negative results in this model, using original-cost values:

• Higher bracket at sale. The deduction saves tax at 24%. Ordinary gain is taxed at 35% after a 3-year hold. Result: $24,782 behind in plain dollars, or $17,796 behind in today's dollars.
• Loss not usable until the sale. 5-year hold, with all the modeled extra deduction usable at sale at 32%. Result: $4,612 behind in plain dollars, or $4,263 behind in today's dollars. You got no early tax saving, and you paid the fee at the start. This setting assumes the required full release of the loss; it does not decide whether you qualify.

What decides whether you come out ahead

1. Your tax rate when you sell compared with when you deducted. A higher sale-year rate can reduce or reverse the benefit. The result also depends on the asset values, fees, and usable deductions.
2. Whether you could use the loss when you took it. If it sat unused until the sale, there was no head start.
3. What the short-life items are really worth at sale. Lower real value means less taxed at your regular rate.
4. How long you hold. An early tax saving has time value. In this example, a longer hold makes that timing benefit larger. It is a deferral benefit, not an actual government loan.

One detail matters in this example. In this hypothetical all-building comparison, the same basis stays in the building: its depreciation saves tax at 32%, and the modeled depreciation-related sale gain is taxed at up to 25%. Moving basis into section 1245 items can turn that later gain into ordinary income instead. Cost segregation can change both timing and rates. That is why the original-cost rows above dip below zero in plain dollars even though the longer holds are positive in today's dollars.

Run your own numbers

The calculator opens with the example above. You can change study or preparer-supplied amounts to explore a hypothetical scenario. It compares with a hypothetical all-building baseline and assumes the same overall sale price and land allocation, enough building gain, and no other gains or losses changing the tax categories. It does not estimate a loss sale, a 1031 exchange, an installment sale, or your tax return. Read its assumptions before using a result.

A worked run: leave everything as it loads and you get $61,673 saved in year 1, $53,976 of extra tax at sale, and $5,585 left in today's dollars. Now change the two tax rates to 24% and 35% and the years to 3. The plain-dollar result drops to $24,782 behind; the today's-dollar result is $17,796 behind.

Other ways out: installment sales, moving in, and holding for life

How you leave the property changes how much of the risk is real.

Does a 1031 exchange avoid cost segregation recapture?

Not always. A study does not block a 1031 exchange. But an exchange does not put off every dollar of recapture by itself. Three things to know:

1. Depreciation labels do not settle 1031 treatment. An item classified as section 1245 property for depreciation can still count as real property under the separate 1031 definition. The IRS Form 8824 instructions explain this. Classify the actual assets; a 5-year or 7-year label alone does not answer it.
2. Personal property does not get real-property exchange treatment. Furniture and ordinary freestanding appliances generally do not qualify as real property for section 1031. A 15% rule for incidental personal property can help with specific identification and qualified-intermediary safe-harbor requirements. For the intermediary safe harbor, the items must be customarily transferred with the real property and total no more than 15% of the replacement real property's value. That does not make the items like-kind or their gain tax-free. The regulation's example receives a $1,000,000 building plus $100,000 of furniture. The furniture passes the incidental-property test, but its receipt still triggers $100,000 of gain in that example. This distinction matters for furnished rentals.
3. Recapture has its own test. Section 1245 recapture depends in part on whether you receive section 1245 replacement property, not simply whether you buy another building or take cash out. IRS Publication 544 gives an exchange with no cash received but $10,000 of ordinary recapture because section 1250 property is received in the exchange. The 2025 Form 8824 instructions address recapture on line 21. Do not treat equal property values or zero cash as a promise of complete deferral.

Also, a tax-deferred exchange is not the fully taxable disposition that generally releases all suspended passive losses. Recognized income may allow some losses to be used, but the exchange does not release them all by itself.

Tell your tax preparer and your exchange company about the study before you list the property. This is much easier to plan before closing than after.

Does cost segregation increase audit risk?

We found no IRS statistic that shows the audit rate for returns with a cost segregation study. The bigger question is what happens if you are examined and the study, or your loss claim, lacks support.

A provider's reported audit rate is not an IRS-measured probability for your return. We do not use those claims to predict your odds.

What we can say from the IRS itself:

• The IRS picks returns in a few ways, including computer scoring and links to other returns it is examining. Being picked does not mean you did something wrong.
• The IRS has a Cost Segregation Audit Techniques Guide, last revised February 6, 2025. It tells examiners what to look for. The guide says on its cover that it is not law and can't be relied on as law.
• That guide does not require one format or one kind of preparer. It asks whether the study is accurate and well documented.

What makes a study hard to defend

The IRS guide lists 13 elements of a quality study. In plain words, an examiner wants to see:

A high price or the word "engineered" does not prove a study has these. And a lower-cost or software study is not bad by default. What counts is whether it fits your property and shows its work.

A real case

In AmeriSouth XXXII, Ltd. v. Commissioner (T.C. Memo. 2012-67), an apartment owner tried to depreciate more than 1,000 building parts over shorter lives, with disputed categories claimed over 5 or 15 years. The IRS denied $1,079,751 of deductions for 2003 to 2005, and the Tax Court upheld most of its arguments. One material detail: the taxpayer stopped participating and did not file its posttrial brief. The court treated otherwise uncontested factual matters as conceded. This was a fact-specific classification and evidence case, not a ruling that all cost segregation fails. (Tax Court opinion, reproduced PDF)

A second case is a warning for buyers. In Peco Foods, Inc. v. Commissioner (T.C. Memo. 2012-18), a buyer had agreed in writing how the purchase price was split among assets. The court would not let the buyer unilaterally change those binding allocations later with a study. The Eleventh Circuit affirmed in 2013, addressing section 1060 and the enforceability of those agreements. If you plan a study, raise it before you sign a contract that spells out the split.

What it costs if a study is cut back

• The extra tax you would have paid without the disallowed deductions
• Interest
• Possibly an accuracy-related penalty of 20% of the underpaid tax. That is 20% of the tax, not 20% of the deduction. It is not automatic.
• Your tax preparer's time to respond

The IRS generally has three years from the return's due date or filing date, whichever is later, to assess more tax, with exceptions. Keep the study, purchase records, and depreciation schedules until that window has closed for the year you dispose of the property. If its basis carries into replacement property in a nontaxable exchange, keep both sets of records until that window closes for the year you dispose of the replacement property. The IRS explains how long to keep property records.

Your study and your loss claim are two different things

A perfect study does not protect a real estate professional claim or a short-term rental claim with nothing behind it. The IRS says you can prove your hours by any reasonable method, such as a calendar or appointment book. Keep one.

What "audit support" really covers

Ask for it in writing. Answering questions about the study, speaking for you to the IRS, testifying, and fixing your return are four different jobs.

Here is one published example. R.E. Cost Seg says on its services page that its audit support is a written response to questions about its engineering method, asset classifications, and study conclusions. It says that support does not include representing you before the IRS, testifying, preparing or amending returns, or defending positions that are not in its study. We read those terms on October 7, 2026. That is the company's published scope, not a test of how it performs.

Ask every provider:

1. Who answers the IRS's questions about the study?
2. What is not included?
3. How long does support last?
4. What costs extra?

We compare how each offer handles documentation, your workload, and audit support.

Find My Cost Seg Provider

No account, email, or phone number needed. The study links on that page are unpaid. How we make money.

Other disadvantages that can change the result

State taxes. Some states do not follow federal bonus depreciation. California is one. Its Franchise Tax Board says the state does not conform to the federal bonus rule. State deductions may be smaller or delayed, and federal and state depreciation schedules can differ. Ask your tax preparer about your state. Don't assume the federal savings repeat on your state return.

The full cost, not just the study fee. Count all of it:

Extra paperwork for property you already own. Applying a study to property placed in service in an earlier year may require an accounting method change and Form 3115. The correct route depends on what was reported and what is changing; some corrections use an amended return instead (IRS Publication 946). A calculation that comes with the study is not the same as someone preparing and filing the form. Ask who does it and what it costs.

An old purchase date stays old. Ordering a study in 2026 does not give a 2022 purchase the 2026 bonus rate. The rate depends on when you acquired the property and put it in service.

Smaller deductions later. You moved the deductions forward. Later years have less.

When to skip, when to wait, and when to compare studies

An open question is not a "no." It is a question to answer first.

If you can use the loss, plan to hold, and your preparer agrees, the next choice is which study fits your property. A lower-cost study only helps if it covers your property and gives your tax preparer what they need.

Find My Cost Seg Provider

You can compare without giving us your name, email, or phone number. The provider handles its own quote and study. Your tax preparer advises on the tax side.

The cost segregation risk worksheet

Fourteen questions: nine for your tax preparer and five for the study provider. Copy it or print it and bring it to both calls.

COST SEGREGATION RISK WORKSHEET
CostSegregationMatch
https://costsegregationmatch.com/cost-segregation-risks/
Tax sources checked October 7, 2026 (UTC).

Get these answers before you pay for a study.
This is a list of questions. It is not tax advice or a tax estimate.

MY NOTES
Property: ______________________________________
Study offer I am looking at: ____________________
How long I plan to keep it: _____________________
My likely exit (sell / 1031 exchange / not sure): _

A. QUESTIONS FOR MY TAX PREPARER

1. Can I use this loss this year? How much can I use now,
   and how much has to wait?
   Answer: _____________________________________

2. How much tax does the usable part save me this year?
   Federal: ______________  State: ______________

3. Does my state follow federal bonus depreciation?
   Answer: _____________________________________

4. If I sell in ____ years, what tax bracket would the sale
   put me in? How much of the gain would be taxed at my
   regular rate instead of the lower gain rates?
   Answer: _____________________________________

5. After the study fee and your fees, how far ahead or behind
   am I compared with not doing a study? What if I sell sooner?
   Planned sale: __________  Earlier sale: __________

6. If I may do a 1031 exchange: which items count as real
   property, what recapture could still be taxed, and what
   should I tell the exchange company before I list?
   Answer: _____________________________________

7. Should I elect out of bonus for a tax class of property?
   That election covers the whole class placed in service
   that tax year. How would it change my deductions?
   Answer: _____________________________________

8. I have owned this property since ______. Do I need
   Form 3115? Who prepares it and who files it?
   Answer: _____________________________________

9. What will you charge to use the study on my return,
   this year and in later years?
   Answer: _____________________________________

B. QUESTIONS FOR THE STUDY PROVIDER

10. Does this offer cover my property type, size, and
    cost basis? Is anything about my property outside it?
    Answer: ____________________________________

11. What facts and records about my property will the report
    use? How do your numbers tie back to what I paid?
    Answer: ____________________________________

12. Can I see a sample report for a property like mine?
    Answer: ____________________________________

13. If the IRS questions the study, what exactly will you do?
    What is not included (speaking for me to the IRS,
    testifying, fixing my return)? How long does support
    last? What costs extra?
    Where this is written: _____________________

14. What is in the price, and what is extra
    (site visit, rush, Form 3115)?
    Total: _____________________________________

C. WHAT IS STILL OPEN

Waiting on my tax preparer: _____________________
Waiting on the provider, in writing: ____________
Waiting on exchange planning: ___________________

NEXT STEP
When the tax side checks out, compare study options:
https://costsegregationmatch.com/find-my-provider/

IRS sources
Loss limits: https://www.irs.gov/publications/p925
Sales and recapture: https://www.irs.gov/publications/p544
1031 exchanges: https://www.irs.gov/instructions/i8824
What examiners look for: https://www.irs.gov/pub/irs-pdf/p5653.pdf

Already got an IRS letter about depreciation?

Do these in order.

1. Read the letter. Find the deadline and the list of what the IRS wants.
2. Call your tax preparer or the person who represents you.
3. Gather the letter, the return, your depreciation schedules, the study, and your purchase or construction records.
4. Ask the study provider for the support its contract promises.
5. Answer the way the letter tells you to, by the deadline.

A new study does not replace a timely response to the letter. The IRS explains each notice type on its own site.

Understand your IRS notice

More questions owners ask

Are software or lower-cost studies riskier?

Not by default. The IRS guide does not require one method. It asks whether the study is accurate and documented. The risk with a do-it-yourself tool is wrong inputs. If you type in the wrong basis or miss a renovation, the report carries the error. Check that the offer covers your property type and size before you pay.

My CPA says not to do it. Who is right?

Ask which risk your CPA means. "It moves deductions earlier" is true, and timing has value. It can also change the tax rate applied to later gain. "Recapture could hurt" depends on the asset values, rates, and sale terms; selling soon is not an automatic loss. "You can't use the loss" deserves a clear answer before you pay. Ask which of these applies, then run the numbers together.

Does a study make sense if I sell in two or three years?

In our original-cost example, a two-year hold is $3,645 behind in plain dollars and $788 behind in today's dollars. Lower component values produce a different result. There is no universal two- or three-year break-even rule. Use the calculator to explore its stated scenario, then have your preparer check the actual sale.

Can I undo a study after I file?

Not simply. The route depends on what was claimed, which returns have been filed, and whether you are correcting an error, changing an accounting method, or changing an election. A later change may need Form 3115, an amended return, or another procedure. Talk to your tax preparer before filing. IRS Publication 946 explains the distinction.

How we checked this page

We checked the primary sources linked beside the claims through October 7, 2026 (UTC), reading the relevant source text beyond search excerpts. The main ones are IRS Publications 925, 544, and 946; the 2025 Form 8824 and Form 4797 instructions; the Cost Segregation Audit Techniques Guide (revised February 6, 2025); Revenue Procedure 2025-32 for the 2026 loss threshold; and the January 14, 2026 bonus-depreciation guidance. We also checked the court opinions. The California example comes from the Franchise Tax Board's 2025 instructions. Provider terms come from the provider's own pages, read during this audit.

The worked example and calculator use our own arithmetic under the stated assumptions. The figures were recalculated and the tools checked in a desktop browser, including mobile-width layouts. They are for learning, not a tax estimate for your property. No tax professional has reviewed this page.

See how we compare study offers. Spot an error? Send a correction and include these details.