Cost Segregation Audit Risk: What Actually Raises Risk

The method is not the risk. The IRS lost in Tax Court almost 30 years ago

Every conversation about cost segregation audit risk should start from the court history. In Hospital Corporation of America v. Commissioner (109 T.C. 21, 1997), the Tax Court ruled against the IRS's position that components inside a building could not be depreciated separately from the building's structural shell. The IRS lost, and it acquiesced to the ruling in 1999. It then published its own Audit Techniques Guide (Pub 5653), a document written for its own examiners describing how a quality cost segregation study is performed and what to look for when reviewing one.

That second fact matters more than owners usually realize. The IRS wrote that guide so its own people could tell a well-built study from a thin one. A study that follows the guide is built the way the IRS describes.

So where does actual audit risk come from

Audit risk sits in execution. Three patterns show up repeatedly in what raises an examiner's eyebrow.

Aggressive percentages without documentation. A study typically shifts about 15 to 35% of a building's basis into faster 5-, 7-, or 15-year schedules, varying meaningfully by property type, restaurants running toward the high end because of dense kitchen equipment, simple shells toward the low end. A study that reports a number well outside that range, with no component-by-component explanation for why, is the kind of thing an examiner is trained to question, regardless of who prepared it.

No engineering behind the classification. The ATG's own quality hierarchy ranks engineering-based methods, where someone actually reviews the building's components, above cost-estimating or questionnaire-based approaches that apply default percentages from a database. A report with no engineer or trained reviewer behind the specific classifications is thinner ground to stand on if a component gets questioned.

No workpapers. A defensible study documents the basis for each item it moves: what it is, why it qualifies for its class life, and how its cost was allocated. A report that hands over a summary number with no supporting detail leaves nothing for anyone, examiner or CPA, to actually review.

The 10-page report problem. The cheap end of this market sells abbreviated studies that run about 10 pages: default percentages by property type, no component-level engineering, no workpapers. Everything above describes that report. It is not that a thin study breaks a rule. It is that when an examiner asks why the carpet, the parking lot, or the kitchen hookups landed where they did, 10 pages have no answer to give. Our reports run 70-plus pages because each classification carries its engineering, and our delivered studies have zero documented IRS audits to date.

What the range looks like in our benchmark runs

The 15 to 35% range is not an abstraction. It shows up across our four commercial benchmark runs, one per property type.

The restaurant sits at the high end of first-year deductions relative to basis because its kitchen carries a dense mix of 5- and 7-year equipment, specific electrical and plumbing runs serving that equipment, that a simple office shell does not have. That is exactly the kind of variation the ATG expects to see, property type driving the percentage, not a firm pushing every study toward the same aggressive number regardless of what the building actually contains. A study where every property type comes back at the same high percentage is itself a signal worth questioning, whether or not any single study in that firm's portfolio is ever examined.

What a proper study actually documents

An engineered study built to the ATG standard classifies a building's reclassifiable components explicitly. Five-year property: carpet, most flooring, decorative lighting, cabinetry, appliances, window treatments, and certain electrical or plumbing serving specific equipment rather than the building generally. Seven-year property: certain fixtures and furniture. Fifteen-year land improvements: paving, fencing, landscaping, site utilities, and outdoor lighting. Everything else stays on the building's structural schedule, 39 years for commercial property, 27.5 years for residential rental, including the roof and central HVAC, which are structural components, not 5-year property, a distinction worth stating plainly because it is a common point of confusion and a common point of overreach in weak studies.

For short-term rentals, the same classification logic applies without a site visit: listing photos from Airbnb or VRBO feed the component review directly, hands-off for the owner but still property-specific rather than a database lookup. Either format, commercial or residential, produces a full engineered report, aligned to the ATG, documenting the basis for every reclassification it makes.

What audit defense of a report actually means

"Audit defense" gets used loosely in this industry, so it is worth being precise about what it covers. Basis provides full audit defense of the report it produces: if an examiner questions a classification or asks for support on a specific component, the engineering team that built the study answers those questions in writing, through the client's CPA. That is defense of the study's technical content.

What audit defense of the report does not mean is representation of the taxpayer. The client's CPA still represents the client in the exam itself, handles the broader return, and manages the relationship with the IRS on everything outside the study's specific technical content. Being precise about that scope matters. A firm that blurs the line between defending its own work and representing a taxpayer in an exam is promising something it is not positioned to deliver. See what audit defense actually covers for the fuller breakdown of that boundary.

How CPA involvement fits into risk management

For a look-back study, on a property owned for years, the reclassification is claimed through Form 3115 with a section 481(a) catch-up deduction in the current year rather than amended returns. The missed depreciation from prior years arrives all at once. Preparing and filing that form is the client's CPA's job. Basis's engineering team takes technical questions directly from the CPA on methodology and classifications, so the person filing the return has a direct line to the people who built the numbers behind it, another piece of the documentation trail.

None of this changes when the property is a purchase, new construction, or a renovation. Land value is excluded first in every case; only the building and its improvements depreciate, and the same classification and documentation standard applies across all three scenarios.

A different risk owners often confuse with this one

Short-term rental owners in particular tend to blend two separate audit questions into one. The first is the question this page addresses: is the cost segregation study itself documented well enough to hold up if a specific component is questioned. The second is entirely different: does the owner's return correctly support treating the rental as non-passive under section 469, so that the accelerated losses a study produces can actually offset other income in the current year rather than being suspended.

That second question runs on its own separate tests. Under Reg. 1.469-1T(e)(3)(ii), a property whose average guest stay is 7 days or less is not treated as a rental activity for section 469 purposes. The owner then needs material participation for losses to be non-passive, most commonly measured by 500-plus hours, substantially all participation, or 100-plus hours and more than any other individual, a test a full-service property manager often breaks since cleaners, co-hosts, and managers all count against the owner in that comparison. Personal use matters too: section 280A limits deductions when personal use exceeds the greater of 14 days or 10% of rental days.

None of these tests change whether the cost segregation study's component classifications are documented correctly. They determine something separate: whether the deductions the study identifies land as non-passive in the year they are claimed. An examiner can question either issue independently, and conflating them, treating a strong average-stay position as if it also proves the study's engineering is sound, or vice versa, misses where each risk actually lives.

What to actually evaluate before choosing a provider

Given where the real risk sits, the questions worth asking a cost segregation provider are about execution. Does an engineer or trained reviewer classify the specific building, or does the report apply default percentages. Does the deliverable include component-level workpapers, or just a summary. Does the reclassification percentage sit inside the normal range for that property type, and if not, is there a documented reason. Does the fee structure create an incentive to inflate the number, see how cost segregation companies price their fees for that comparison. And what, precisely, does the firm's audit support actually cover.

These are the same questions worth asking about Basis specifically, and the honest answer to each is in the spokes below: what the IRS's own guide looks for, what actually happens if a study gets questioned, whether an engineer is required, where DIY approaches genuinely fit, the precise scope of audit defense, and the red flags worth watching for from the buyer's side, including in firms that are not Basis.

Documentation matters past the exam, too

A study's workpapers do not stop mattering once the return is filed and no exam ever comes. They resurface at sale. Gain attributable to depreciation on the 5- and 7-year property a study reclassifies, 1245 property, is recaptured at ordinary rates when the property sells, while straight-line depreciation on the structural 39-year or 27.5-year real property is unrecaptured section 1250 gain, taxed up to 25%. Calculating that recapture correctly depends on knowing exactly what was reclassified and when, the same component-level detail the ATG asks for during an exam. A 1031 exchange can defer both categories of gain, including on a property carrying a prior cost segregation study, when the replacement property rules are met, but that deferral calculation still runs on the original study's documentation.

See what documentation a study should include and what audit experience actually looks like in practice for more on both ends of this timeline, filing and eventual sale.

Getting a study built to hold up, not just to sound impressive

A free Preliminary Benefit Estimate at /qualify models a property's likely first-year acceleration before any commitment, using the same engineering-based approach that produces the documentation trail described above, not a database percentage.

Whether a particular study's classifications fit your property's facts, and how any given exam outcome would apply to your return, are questions for your CPA. What a properly built study and its workpapers show is the number the mechanics produce for that specific building, documented the way the IRS's own guide says it should be.

Every guide in this series

Frequently asked questions

Does claiming cost segregation increase my audit chances?

There is no published IRS data establishing that a cost segregation study by itself increases exam selection odds. What raises scrutiny within a study that gets examined is unsupported reclassification percentages, missing documentation, or a non-engineered methodology on a complex property, not the act of claiming the deduction.

Is cost segregation a gray area or a loophole?

The IRS lost Hospital Corporation of America v. Commissioner in 1997, and it publishes its own Audit Techniques Guide describing how a proper study should be done. A study that follows that guide is executing a method the IRS itself describes.

What happens during a cost segregation audit?

An examiner reviewing a cost segregation claim generally requests the study's supporting workpapers and may ask about specific component classifications. A well-documented, engineering-based study answers those questions directly with component-level support. See the full walkthrough of what happens if a study is questioned for more detail.

Who defends the study if the IRS asks questions?

Basis provides full audit defense of the report it produces, meaning the engineering team that built the classifications answers an examiner's technical questions about the study itself, in writing, through the client's CPA. The client's CPA continues to represent the client in the broader exam.

Keep reading