The Cost Segregation Audit Defense Playbook

What an IRS examination of cost segregation actually looks like

Cost segregation examinations are not adversarial. The IRS publishes the Cost Segregation Audit Techniques Guide (ATG) — a multi-chapter publicly-available document — precisely so taxpayers and practitioners know what to expect. The examination follows a defined sequence.

Initial request: The examining agent requests the engineering report and supporting workpapers for the property under examination. The request typically arrives 6–18 months after the tax return claiming the cost segregation deduction is filed.

Document review: The agent reviews the report against the ATG framework. The ATG explicitly identifies the elements of a “quality” cost segregation study:

1. Preparation by an individual with construction or engineering expertise
2. Detailed engineering report with methodology disclosure
3. Use of recognized cost data (e.g., RSMeans)
4. Site visit (for commercial properties)
5. Component-level identification with documentation
6. Treatment of components in accordance with current tax law

The first IRS step is to verify these elements are present. Studies that lack any of them — particularly the site visit and component-level documentation — face escalating scrutiny.

Sample examination: If the report passes the initial review, the agent typically selects a sample of the highest-dollar component classifications for detailed examination. “Why is this $42,000 of HVAC equipment classified as 5-year personal property rather than 39-year structural?” The engineering documentation in the report must answer the question.

Resolution: Most examinations resolve at one of three outcomes:

• No change — the documentation supports the classifications as filed. The examination closes without adjustment.
• Partial recharacterization — some specific components are recharacterized (e.g., a particular HVAC zone is determined to be base building rather than tenant-specific). The depreciation schedule is adjusted; the broader cost segregation position is preserved.
• Full recharacterization — rare. Generally only seen when the underlying study failed to apply the engineering methodology at all (no site visit, no component documentation, blanket percentages applied).

What documentation defends each classification

A defensible cost segregation report documents each component classification at three levels:

Level 1: The component itself

Every reclassifiable component must be inventoried with sufficient specificity to identify it. The component schedule lists:

• Component description (e.g., “Dedicated condensing unit, 5-ton, Tenant Suite 200, second floor”)
• Quantity and unit
• Unit cost (with source citation, typically RSMeans 2024 or comparable)
• Total cost
• MACRS classification (5-year, 7-year, 15-year, or 39-year)
• Section reference (1245 or 1250)

Studies that omit any of these fields invite line-by-line scrutiny. Studies that include all of them give the examining agent enough information to reach the same conclusion without challenging the underlying engineering.

Level 2: The engineering basis

For any component whose classification requires defense — typically items at the boundary between personal property and structural — the report includes the engineering rationale. The five most common boundary cases:

Dedicated HVAC vs base building HVAC. The classification turns on whether the equipment serves a specific tenant suite or the building as a whole. The engineering documentation includes a photograph of the equipment, a schematic showing duct runs and zone boundaries, and the circuit schedule. Treasury Regulation 1.48-1(c) controls the test.

Decorative vs functional lighting. Pendant lights, sconces, accent lighting, and track lighting installed for aesthetic effect classify as 5-year. Recessed cans providing general illumination as base electrical are 39-year. The lighting plan and circuit schedule resolve the split. Photographs of the installation establish the type.

Movable vs permanent interior partitions. The Section 1.48-1(c) permanence test: partitions that can be removed without damage to the structural envelope qualify as 5-year. Permanent walls structural to the building do not. Mounting details (top track, floor track, connection to demising walls) are documented photographically.

Tenant improvements paid by landlord. TIs that meet the permanence test, classified into accelerated buckets. The lease and the capital improvement records establish the landlord’s ownership of the improvement.

Site work. Parking lots, sidewalks, monument signage, and landscaping are 15-year land improvements under Section 168(e)(3)(E). The audit looks at the cost allocation against the site work line items in the original construction documents or acquisition closing settlement.

Level 3: The methodology

The report’s methodology section documents how the engineer arrived at each classification. The narrative cites:

• The Treasury Regulation, IRS Revenue Procedure, or judicial authority controlling the classification
• The specific facts of the property that fit within the cited authority
• Cross-references to the photographs, schematics, or schedules in the appendices

A methodology narrative that says “I used my professional judgment” is not defensible. A methodology narrative that says “Pursuant to Treasury Regulation 1.48-1(c), this dedicated condensing unit serving Tenant Suite 200 is properly classified as 5-year personal property because it is removable without damage to the structural building envelope, as shown in the mounting detail at Appendix B-12” is defensible.

What separates engineered studies from “modeling” studies

The IRS ATG explicitly distinguishes between engineered studies (with site visit, component identification, and cost allocation) and “modeling” or “rule-of-thumb” approaches (industry-standard percentages applied without underlying engineering).

Modeling studies are accepted in some contexts — typically very small properties under $500K basis where the cost of a full engineering study outweighs the marginal benefit of the additional reclassification. For commercial properties at meaningful basis levels, modeling-only studies face higher examination risk.

The practical difference at audit:

• An engineered study with site visit, component schedule, and methodology narrative: examination typically resolves at “no change” or “minor partial recharacterization on specific components.”
• A modeling study applying industry percentages: examination may recharacterize broad categories of components, especially when the percentages differ materially from the property’s actual component density.

The latter is why “what gets sold cheap” in cost segregation matters at examination. A $1,000 cost segregation “study” that’s really an Excel template with industry percentages applied won’t defend a $400K accelerated bucket. The engineering work the IRS expects to see won’t be in the file.

What audit defense from your cost segregation provider should look like

At Commercial Cost Seg, audit defense is included in the engagement scope for the life of the study. If the property’s depreciation is examined, the engineering team that prepared the study supports the position at no additional cost. This is standard practice among engineering-grade cost segregation firms.

Three things distinguish good audit defense from bad:

Same engineer. The engineer who prepared the study, who visited the property, who classified the components, is the engineer who defends the study at audit. Handing the defense to a different person who has to re-learn the property at the engagement of an active examination is how partial recharacterizations happen.

Documentation is in the report. A study where audit defense requires producing additional workpapers after the fact is a study that should have produced those workpapers up front. The 30+ page engineered report contains everything the examining agent will request.

Recharacterization-resistant classifications. Engineering judgment that anticipates the IRS line of questioning at preparation produces classifications that survive scrutiny. Aggressive classifications that depend on technicalities or boundary interpretations are the ones that get partially recharacterized.

When to expect an examination

The IRS examines cost segregation on a small percentage of cost-segregated properties. The selection is driven by total deduction size, taxpayer profile, and broader exam-cycle factors — not by the cost segregation election itself.

Most examinations occur 12–24 months after the return claiming the deduction is filed. Section 481(a) catch-up deductions filed via Form 3115 are sometimes selected at higher rates because of the deduction concentration in a single year. The catch-up itself is not the trigger; the deduction size is.

Bottom line for commercial owners

Cost segregation is a strategy the IRS expects to see and has published a roadmap for examining. The strategy’s defensibility at audit depends almost entirely on the engineering documentation behind it. Engineered studies with site visits, second-engineer review, component schedules, and methodology narratives generally hold up. Modeling studies and template-driven studies face escalating recharacterization risk.

For property owners and CPAs evaluating cost segregation, the engineering quality of the provider is the most consequential decision — more so than the fee, more so than the turnaround time, more so than the format of the deliverable.

See our methodology in detail → or schedule a scoping call to discuss audit defense for a specific property.