Cost segregation and bonus depreciation: why the combination does the work

A cost segregation study and bonus depreciation are two halves of one machine. The study finds the 5-, 7-, and 15-year property inside a building's purchase price; Section 168(k) then lets the owner deduct 100 percent of that property's cost in year one — for qualified property acquired after January 19, 2025, permanently. Without the study, bonus has nothing in a building to attach to, because 27.5- and 39-year real property is never bonus-eligible. Without bonus, the study still accelerates deductions, but over 5 to 15 years rather than immediately.

The One Big Beautiful Bill Act, enacted July 4, 2025, ended the TCJA phase-down and restored the 100 percent rate, which is why cost segregation activity has surged again. This piece explains the statutory interaction, the transition rules that still govern 2023–early-2025 acquisitions, the special role of qualified improvement property, and — with a worked example — what the year-one numbers actually look like.

What Section 168(k) actually covers

Bonus depreciation — formally the "additional first-year depreciation deduction" — applies to qualified property: MACRS property with a recovery period of 20 years or less, certain computer software, water utility property, and qualified improvement property. The building itself, at 27.5 or 39 years, is categorically out. That single line of the statute is the entire reason cost segregation and bonus travel together: the study converts a slice of 39-year property into 5-, 7-, and 15-year property, and every dollar it converts becomes bonus-eligible. (For the study mechanics themselves, start with what a cost segregation study is.)

Two TCJA-era features remain important:

• Used property qualifies. Before 2018, bonus applied only to original-use property. Section 168(k)(2)(E)(ii) now extends it to used property on the taxpayer's first use, provided it is not acquired from a related party. Buying a 30-year-old strip center and running a study produces the same bonus eligibility as building new.
• Bonus is automatic unless you elect out. The election out under Section 168(k)(7) is made by class of property, per year. Electing out of the 15-year class while taking bonus on 5-year property, for instance, is a legitimate income-smoothing tool for taxpayers who cannot absorb the full deduction.

The rate: 100 percent again, with a seam

The OBBBA made 100 percent bonus depreciation permanent for qualified property acquired after January 19, 2025 and placed in service after that date. Property acquired on or before that date remains governed by the TCJA phase-down. The seam matters for anyone doing a look-back or catching up on recent acquisitions:

Bonus rates by acquisition and placed-in-service timing:

"Acquired" carries the Section 168(k) meaning developed under the TCJA regulations: generally when a written binding contract is entered, or for self-constructed property, when physical work of a significant nature begins. A building under a binding contract signed January 10, 2025 and closed in June 2025 is a 40 percent asset, not a 100 percent one. Examiners know to check contract dates; so should the preparer.

Qualified improvement property: the 39-year exception

One category of real property does get bonus: qualified improvement property — generally any improvement to the interior of a nonresidential building placed in service after the building itself, excluding enlargements, elevators and escalators, and internal structural framework. After the CARES Act retroactively fixed the TCJA's drafting error, QIP is 15-year property (straight line) and therefore bonus-eligible.

For renovation-heavy fact patterns, QIP can dwarf the traditional 5- and 15-year reclassifications: an interior build-out that is neither structural nor an enlargement may be 100 percent deductible in year one even before the study finds a single dedicated circuit. A study on a renovation typically does both jobs — segregating true Section 1245 property and documenting which remaining costs qualify as QIP. The full treatment is in our brief on qualified improvement property.

A worked example

Assume a taxpayer acquires a medical office building in March 2026 for $4,000,000, of which $800,000 is allocated to land, leaving $3,200,000 of depreciable basis. A study reclassifies 25 percent to 5-year property and 10 percent to 15-year land improvements. Mid-month convention applies to the 39-year property (9.5 months of a 39-year straight line in year one); the short-life property is acquired after January 19, 2025, so 100 percent bonus applies.

Year-one depreciation, with and without a cost segregation study:

At a 37 percent marginal rate, the incremental first-year deduction of roughly $1.1 million is worth about $406,000 of deferred tax — against a study fee that might run $8,000 to $15,000 for a property of this size. The same building acquired under a December 2024 binding contract and placed in service in 2025 would get 40 percent bonus on the short-life property, cutting the year-one figure to roughly $600,000 (bonus plus regular MACRS on the remainder) — still substantial, but a different conversation.

Two caveats keep the example honest. First, this is deferral: the taxpayer has simply used up future deductions, and on sale the accelerated portion is recaptured at ordinary rates under Section 1245 — see depreciation recapture explained. Second, the deduction is only worth $406,000 if the taxpayer can use it in 2026.

Can the loss actually be used?

Bonus-driven cost segregation routinely produces a rental loss far exceeding the property's income. Whether that loss reaches the owner's other income depends on a gauntlet:

• Section 469 passive activity rules. Rental losses are per se passive unless the owner qualifies as a real estate professional and materially participates (or fits the short-term-rental fact pattern discussed in our STR brief). Passive losses offset only passive income; the excess is suspended.
• Section 461(l). Even non-passive business losses are capped ($313,000/$626,000 as indexed for 2025, adjusted annually); the excess becomes an NOL carryforward.
• Basis and at-risk limits for partners and S corporation shareholders.

A seven-figure deduction that lands in a suspension account has a present value of roughly zero until it is released. This is the single most common way a technically flawless study fails economically, and it is why the feasibility analysis belongs with the CPA, not the study provider. The full anti-pitch is in when cost segregation doesn't make sense.

Electing out, and other dials

100 percent bonus is a default, not a mandate, and taking it is not always optimal:

• Rate arbitrage in the wrong direction. A taxpayer temporarily in a low bracket — a start-up year, a large NOL — may prefer to elect out by class and preserve deductions for higher-rate years.
• Section 163(j) interaction. Depreciation is no longer added back in computing adjusted taxable income (the EBIT-based limit has applied since 2022, and OBBBA's restoration of an EBITDA-style addback for years after 2024 changes this again for many taxpayers — check the current computation), so large depreciation deductions can constrain interest deductibility for electing real property trades or businesses and everyone else differently. Electing real property businesses that took the Section 163(j)(7) election must use ADS for their real property — which kills bonus on 15-year land improvements and QIP but not on 5- and 7-year personal property.
• State conformity. Many states decouple from Section 168(k) entirely or in part. A study that produces a large federal deduction may produce a large state addback and a multi-decade state depreciation schedule to track.

The mechanics of depreciation elections and conventions generally are covered in depreciation basics.

The bottom line

Bonus depreciation is the multiplier on everything a cost segregation study finds. At 100 percent, the study's reclassified basis becomes a year-one deduction; at the phase-down rates that still govern 2023–early-2025 acquisitions, it becomes a partial one. The statutory interaction is clean. The planning question — whether the taxpayer can absorb the deduction, tolerate the recapture, and navigate state decoupling — is not, and it deserves as much attention as the headline number.