← Small-Business Tax Research Benchmark
Deduction · IRC §168 Typical recovery $25K to $200K one-time

Cost segregation: recovering hidden depreciation on commercial real estate

Commercial real estate depreciates over 39 years by default. Residential rental over 27.5 years. Inside almost every building, 15% to 35% of the basis is actually 5, 7, or 15 year property under the §168 class life rules, and that portion can be depreciated dramatically faster. A cost segregation study reclassifies it; on older buildings, a §481(a) catch-up via Form 3115 recovers the deferred deductions in the current year.

TL;DR

A cost segregation study breaks a building's basis into property classes with different recovery periods under IRC §168. Land improvements (parking lots, landscaping, fencing) are 15 year property. Personal-property items physically attached but not structurally integral to the building (specialty electrical, decorative finishes, certain mechanicals, removable partitions) are 5 or 7 year property. The rest stays at 27.5 or 39 years. Faster recovery periods mean larger deductions in early ownership years; combined with bonus depreciation (40% for property placed in service in 2025 under OBBBA 2025, scheduled to phase down further) the front-loaded deduction is meaningful. For buildings placed in service in prior years, Form 3115 §481(a) lets the taxpayer catch up the entire prior-period understated depreciation in the year of the change of accounting method, without amending prior returns. Typical one-time recovery range: $25K to $200K depending on basis, building type, and how long ago the building was placed in service.

Why the deferred deduction exists

The default convention when a commercial building is placed in service is to depreciate the entire basis (less land) over 39 years on a straight-line basis. The default convention is wrong for any building that contains more than the bare structural shell, because §168 and the underlying class life regulations recognize that buildings contain personal property and land improvements that should not be on the 39-year clock.

The 1997 Tax Court decision in Hospital Corporation of America established that a study using engineering principles can support reclassification of building components into shorter recovery periods. The IRS responded in 2004 with Revenue Procedure 2004-11 (the framework for changing accounting methods via Form 3115) and later with the Cost Segregation Audit Technique Guide, most recently updated in February 2025. The ATG sets the substantiation standard the IRS expects: an engineering-based methodology with a defensible component-level allocation.

For new construction or acquisition, the cost segregation study is run in the year of placement in service and applied on the original return. For prior-year buildings, the §481(a) catch-up via Form 3115 lets the taxpayer recover the entire deferred-depreciation amount in the current year as a single deduction, without amending each prior return.

Who benefits most

Cost segregation is a real-estate-owner tool. The dollar leverage scales with building basis, with the percentage of basis that can be reclassified, and with the marginal federal-plus-state tax rate of the owner. A short list of fact patterns where the study reliably pays for itself many times over:

  • Commercial buildings with basis above $500K, including retail, office, hospitality, industrial, manufacturing, medical-office, and self-storage. The component density and the engineering integral-vs-decorative split typically support 20% to 35% reclassification into shorter-life property.
  • Restaurants and hospitality. High concentration of specialty electrical, plumbing, decorative finishes, and equipment-related building improvements. Reclassification commonly hits 30% to 40% of basis.
  • Multifamily and residential rental above $1M basis. Carpets, decorative lighting, appliances, certain millwork, and land improvements are reclass candidates from the 27.5-year baseline.
  • Manufacturing and warehousing. Specialty utility infrastructure tied to equipment (process piping, dedicated electrical, compressed air, dust collection) reclassifies cleanly when it serves the operating function rather than the building.
  • Buildings placed in service in the last 15 years that have not been studied. The §481(a) catch-up captures the entire deferred-depreciation balance in the current year; the older the building, the larger the catch-up.

Owners who plan to hold the building long enough to outrun the depreciation recapture see the cleanest economics. For a building expected to be sold within 5 years, run the recapture math first; cost segregation accelerates the timing of the deduction, it does not change the total, and the unrecovered §1245 personal-property depreciation comes back at ordinary rates on sale.

Why the engineering-based methodology matters

The IRS Cost Segregation Audit Technique Guide (02/2025) is clear about what it expects to see: a study performed by a qualified individual using engineering principles, with a component-level breakdown documented to the level of the underlying construction draws or the asset appraisal.

Three methodology tiers, in descending order of audit defensibility:

  1. Detailed engineering approach (from construction draws). Component-by-component reclassification based on the actual construction-draw schedule. Highest defensibility, available when the property was built or substantially renovated under the current owner.
  2. Detailed engineering approach (from physical observation). Site walk by a qualified engineer, sampling of representative components, and allocation based on industry-standard percentages tied to the underlying construction draws. Strong defensibility for acquired buildings or older holdings.
  3. Survey or sampling approach. Allocation based on industry-standard percentages without component-level engineering. The IRS ATG specifically flags this as the most vulnerable approach and the one most likely to lose at exam. Studies that price themselves at a fraction of the engineering-based fee are typically using this method; the discount is real, the risk is real.

For any building with basis above $1M, the engineering-based methodology is the only credible choice; the fee differential pales next to the reclassification difference, and the audit-defensibility delta is meaningful. Below $1M basis, the survey approach can still work if the qualified individual is genuinely qualified and the sampling is documented; check the deliverable for component-level allocation, not just summary percentages.

A worked example

Owner-operator acquired a $3,200,000 retail building in 2020 (basis less land allocation: $2,800,000). No cost segregation study was ever performed. Standard depreciation has been computed on the full $2,800,000 over 39 years, generating about $71,800 of annual depreciation, or roughly $359,000 of cumulative depreciation through end of 2024.

A 2026 engineering-based cost segregation study finds 22% of basis reclassifies into 5 and 15 year property:

  • 5-year property (specialty electrical, decorative finishes, certain mechanicals): 14%, $392,000
  • 15-year property (land improvements, parking lots, landscaping, fencing): 8%, $224,000
  • 39-year property (remaining structural shell, building systems): 78%, $2,184,000

The reclassified property, depreciated under MACRS from the 2020 placement-in-service date with applicable bonus depreciation, would have generated substantially more cumulative depreciation through 2025 than the straight-line 39-year method. The §481(a) catch-up captures that difference as a single deduction in 2026. At an owner's combined federal-plus-state rate of 35%, the catch-up commonly translates to a federal tax recovery in the $80,000 to $150,000 range on a building of this size, depending on the year placed in service and the applicable bonus-depreciation percentages.

The catch-up is taken on Form 3115 (Application for Change in Accounting Method), filed under the automatic-consent procedures in Revenue Procedure 2024-23 (the most recent auto-consent revenue procedure). No private-letter ruling is required for the standard reclassification; the automatic-consent procedures cover it as DCN 7.

Common mistakes that lose money or invite an exam

  1. Picking a non-engineering provider for a >$1M building. The fee discount is the bait. The IRS ATG specifically calls out non-engineering studies as the highest-exam-risk category. If the provider does not have a P.E. or a qualified individual on the team and is not pulling component-level data from construction draws or a physical walkthrough, the study is exam bait.
  2. Over-aggressive personal-property classification. Building components that are structurally integral to the operation of the building (load-bearing electrical, primary HVAC distribution, plumbing systems serving the building generally) are §1250 real property regardless of how they look on a draw schedule. Studies that pull these into 5-year personal property lose at exam.
  3. Missing the §481(a) catch-up window. Form 3115 for the automatic-consent cost-segregation change is filed with the current-year return; there is no separate amendment cycle. Wait until after filing and you cannot retroactively pick up the catch-up; the change has to be made on a timely-filed return for the year of the change.
  4. Forgetting the recapture math on planned sales. §1245 personal-property depreciation recaptures at ordinary income rates on sale, not capital-gains rates. Owners planning to sell within 3 to 5 years should run the math both ways before commissioning the study; on short holds the deferral can be flat or negative once recapture is included.
  5. Not coordinating with the §179D commercial energy-efficient building deduction. If the building had qualifying energy-efficient lighting, HVAC, or envelope upgrades, §179D can stack with the cost segregation reclassification; the two analyses are typically done together. See /insights/section-179d-energy-efficient-buildings.

Why generalist preparers do not initiate the study

Most CPAs running 300 returns a year do not have the engineering capacity to perform a defensible cost segregation study in-house. The standard practice is to refer the work to a specialist provider and bill the integration on the return side. Two reasons that handoff does not happen as often as it should:

  1. The owner does not know the option exists. Real-estate operators with one or two properties have not had cost segregation come up unless their CPA proactively raised it, and most do not. The owner sees depreciation as a fixed schedule on the K-1 and does not realize a different schedule is on the table.
  2. The CPA does not want to vet the specialist provider. Cost segregation has a reputable specialist tier and a problematic discount tier. CPAs who have not built relationships with the reputable specialists do not refer at all rather than refer to the wrong one.

The case for a feasibility check on any commercial building with basis above $500K, even when nobody has raised it, is straightforward: the feasibility analysis itself is cheap (most reputable providers run a free or low-fee estimate based on the basis, asset type, and year placed in service) and either returns a defensible recovery estimate or it does not.

Related reading

Source authority: IRS Audit Technique Guide, Cost Segregation (02/2025). Current ATG library at irs.gov/businesses/audit-techniques-guides-atgs.

Have a cost-seg question? Get free research.

Send us the building basics (type, basis, year placed in service, prior depreciation method) and we will tell you, in writing with statute citations, whether a study makes sense and what the likely §481(a) catch-up looks like. Five business days, no sales pitch, no contract gate.

Send your question →