Cost Segregation Statistics (2026): The 20–40% Rule, Checked
By CostSegregationMatch · Updated October 2026
In 13 published commercial building reports, the median share moved to 5- or 15-year property was 31.3% of depreciable basis (cost, not counting land), with a range of 20.0% to 57.3%, according to CostSegregationMatch's October 2026 analysis of Engineered Tax Services' case reports. Four landed above 40%.
Key cost segregation statistics
31.3% median. Across 13 commercial building case reports that Engineered Tax Services published in 2026, the median share of depreciable basis assigned to 5- or 15-year property was 31.3% (CostSegregationMatch analysis, October 2026).
20.0% to 57.3% range. Across 13 commercial building reports that Engineered Tax Services published in 2026, the share of depreciable basis assigned to 5- or 15-year property ran from 20.0% for a gym in Joplin, Missouri, to 57.3% for a fast-food restaurant in Manning, South Carolina (CostSegregationMatch analysis, October 2026).
9 inside the rule, 4 above it. Nine of 13 commercial building reports that Engineered Tax Services published in 2026 assigned 20% to 40% of depreciable basis to 5- or 15-year property. Four were above 40%. None was below 20% (CostSegregationMatch analysis, October 2026).
34.0% mean, 27.9% by dollars. Across 13 commercial building reports that Engineered Tax Services published in 2026, the mean share of depreciable basis assigned to 5- or 15-year property was 34.0%. Counted by dollars, $15.8 million of $56.8 million in depreciable basis, or 27.9%, went to those classes (CostSegregationMatch analysis, October 2026).
Published benchmarks disagree. In 5 of the 9 building types displayed here, at least two published "typical" reclassification ranges do not overlap, across seven benchmark sources checked in October 2026 (CostSegregationMatch analysis).
Apartments and rental houses. The median low and high ends of the published "typical" ranges are 20% and 35% for multifamily buildings (six sources) and 20% and 34% for single-family rentals (five sources). These summarize published ranges, not measured study results (CostSegregationMatch analysis, October 2026).
Mobile home parks: 97.6% to 100.0%. Four mobile home park case reports published by Engineered Tax Services in 2026 assigned 97.6% to 100.0% of depreciable basis to property with lives of 15 years or less (CostSegregationMatch analysis, October 2026).
$19,510 per $100,000 in our model. Moving $100,000 from a 39-year schedule to a full first-year write-off is worth about $19,510 in present value, assuming July placement, a 37% federal tax rate, a 5% discount rate, full use of deductions and no sale. For a residential rental it is about $15,874 (CostSegregationMatch calculation from IRS Publication 946 tables, October 2026).
About half of the first-year tax cut. In our July commercial-building model, moving $100,000 to 100% first-year bonus gives a $19,510 present-value gain, 53% of the $36,565 first-year federal tax cut. It assumes a 37% federal rate, 5% discount rate, full use of deductions and no sale (CostSegregationMatch calculation, October 2026).
$308,239 vs. $11,770. On $1 million of commercial depreciable basis, excluding land, with 30% reclassified, the modeled first-year depreciation deduction is $308,239 with a study and $11,770 without, for property placed in service in July under 100% bonus depreciation (CostSegregationMatch calculation from IRS Publication 946 tables, October 2026).
Twice as much at 100% bonus. On $1 million of commercial depreciable basis with 30% reclassified, our July model gives a first-year deduction of $308,239 at 100% bonus depreciation and $155,239 at 40%, about twice as much. The reclassified share is split two-thirds 5-year and one-third 15-year property, using the half-year rule (CostSegregationMatch calculation from IRS Publication 946 tables, October 2026).
100% bonus depreciation. Qualified property acquired and placed in service after January 19, 2025 can get a 100% first-year allowance (IRS Publication 946, 2025 edition; IRS Notice 2026-11).
$495 software example; $3,000 to $12,000 provider range. CostSegregation.com displayed a $495 software price at $750,000 of residential tax basis, excluding land. Engineered Tax Services says most professionally prepared studies run $3,000 to $12,000, depending on property size and complexity (provider pages checked October 6, 2026).
13 elements. The IRS guide its examiners use to review these studies lists 13 principal elements of a quality study (IRS Publication 5653, February 2025 revision, checked October 6, 2026).
On this page: What the case reports show · The 20–40% rule · Parks and farms · What it is worth · The 2025 law · Study prices · The IRS · Limits · Method · Cite · Data · FAQ
What percentage of a building does a cost segregation study reclassify?
In 13 commercial building case reports that Engineered Tax Services published in 2026, the median share assigned to 5- or 15-year property was 31.3% of depreciable basis. The lowest was 20.0% and the highest was 57.3%. These are examples one provider chose to publish, not a national average.
First, the words. A commercial building is normally written off over 39 years, and a residential rental over 27.5 years. A cost segregation study pulls out the parts that the tax rules let you write off faster. 5-year property includes appliances and carpet used in residential rentals. 15-year property can include depreciable landscaping and pole-mounted parking-lot lights. The right class depends on the asset and how it is used. The examples come from IRS Publication 946 and the IRS cost segregation guide. Depreciable basis is the cost you are allowed to depreciate. It leaves out land, because land does not wear out.
So "31.3%" means this: out of every $100 of building cost (not land), $31.30 moved into the fast lanes.
| Case report | Depreciable basis | 5-year property | 15-year property | Share in 5- or 15-year property |
|---|---|---|---|---|
| Gym, Joplin, MO | $1,380,000.00 | $206,237.92 | $70,374.80 | 20.0% |
| Retail pharmacy, Turlock, CA | $600,208.24 | $62,849.44 | $67,819.88 | 21.8% |
| Warehouse / office, Joliet, IL | $2,100,000.00 | $389,455.50 | $84,869.67 | 22.6% |
| Warehouse / manufacturing, Corona, CA | $37,471,308.62 | $5,168,110.56 | $3,816,220.84 | 24.0% |
| Warehouse / office, Elmhurst, IL | $2,300,000.00 | $385,569.16 | $226,749.02 | 26.6% |
| Office, Tampa, FL | $1,677,520.00 | $354,854.94 | $136,393.84 | 29.3% |
| Auto service facility, Richardson, TX | $1,031,510.00 | $214,505.61 | $108,750.92 | 31.3% (median) |
| Medical office, Floyds Knobs, IN | $1,718,170.00 | $427,528.95 | $204,270.07 | 36.8% |
| Preschool, Wesley Chapel, FL | $3,020,340.93 | $497,915.00 | $680,522.00 | 39.0% |
| Preschool, Cape Coral, FL | $1,215,785.00 | $230,670.00 | $261,746.00 | 40.5% |
| Dental office, Belmont, NH | $691,841.50 | $223,473.22 | $91,241.76 | 45.5% |
| Warehouse / storage, Katy, TX | $1,260,360.00 | $159,826.80 | $429,008.62 | 46.7% |
| Fast-food restaurant, Manning, SC | $2,292,636.00 | $664,605.32 | $649,068.16 | 57.3% |
Source: CostSegregationMatch analysis of Engineered Tax Services' published case reports. Each report name links to its source page. All pages checked October 6, 2026. Share = (5-year dollars + 15-year dollars) ÷ depreciable basis.
How we got 31.3%
A median is the middle number. Line up the 13 shares from smallest to largest and the seventh one is the middle. Here that is the auto service building in Richardson, Texas: $323,256.53 ÷ $1,031,510.00 = 31.3%.
Six reports sit below it and six sit above it. One very high or very low report cannot drag a median around, which is why we lead with it.
What if you average them instead?
You get three honest answers, and they differ.
| Measure | Result | What it means |
|---|---|---|
| Median | 31.3% | The middle report |
| Mean (simple average) | 34.0% | Add the 13 shares, divide by 13 |
| Dollar-weighted share | 27.9% | $15,812,638 ÷ $56,759,680.29 |
Source: CostSegregationMatch analysis of Engineered Tax Services' published case reports, checked October 6, 2026.
The mean is higher because the Manning restaurant, at 57.3%, pulls it up. The dollar-weighted share is lower because one giant building dominates the dollars. The Corona warehouse has $37.5 million of the $56.8 million in basis, and it came in at 24.0%.
If you are quoting one number, quote the median and say "13 published commercial building reports" with it.
Where did the fast dollars go?
Of the $15.8 million that moved, 56.8% went to 5-year property and 43.2% went to 15-year property. In four of the 13 reports, 15-year property was the bigger piece. The Katy warehouse and storage building is the clearest case: 34.0% of its basis was 15-year property and only 12.7% was 5-year.
Do building types differ?
They do, but the groups here are tiny, so treat these as examples.
- Warehouse-related buildings (4 reports): 22.6%, 24.0%, 26.6% and 46.7%. The median is 25.3%.
- Office, medical and dental (3 reports): 29.3%, 36.8% and 45.5%.
- Preschools (2 reports): 39.0% and 40.5%.
Source: CostSegregationMatch analysis of Engineered Tax Services' published case reports, checked October 6, 2026.
Is the 20% to 40% rule a real average?
KBKG and Miller Cooper both publish 20% to 40% as an average. Neither page gives the sample or calculation behind that stated average. In the 13 case reports above, nine landed inside that range and four landed above it.
In this sample, the rule covers nine of 13 reports, but it is not a ceiling.
Cost segregation studies are usually said to move 20% to 40% of a building's cost into faster write-offs.
Here is who says what:
- KBKG says that on average 20% to 40% of a property's components can be written off faster (page checked October 6, 2026).
- Miller Cooper says the average study moves 20% to 40% of depreciable costs (checked October 6, 2026).
- Engineered Tax Services says its studies typically reclassify 20% to 45% of cost basis (checked October 6, 2026). Three of its 13 case reports here landed above 45%.
- R.E. Cost Seg's article says most properties reclassify 25% to 45% of depreciable basis (checked October 6, 2026).
Several firms go further and publish a "typical" range for each building type. We put ranges from seven of those sources side by side.
| Building type | HCVT | Miller Cooper | Engineered Tax Services | Overline | Modern CFO | Cost Seg Smart | AE Tax Advisors |
|---|---|---|---|---|---|---|---|
| Warehouse | 10–30 | 12–20 | — | 15–28 | 15–28 | 15–28 | 10–20 |
| Office | 10–40 | 12–30 | — | 26–38 | 20–32 | 16–29 | 15–25 |
| Retail strip | 20–40 | 12–30 | — | 30–38 | 22–34 | 20–37 | 20–30 |
| Multifamily | 20–50 | — | 20–40 | 28–38 | 22–32 | 14–26 | 15–25 |
| Single-family rental | 20–40 | — | 20–40 | 24–34 | 24–34 | 9–32 | — |
| Hotel | 25–45 | — | 22–45 | 32–42 | 26–38 | — | 22–35 |
| Restaurant | 20–50 | — | 25–45 | 32–44 | 28–42 | 16–29 | 25–40 |
| Medical office | — | — | 20–40 | 30–38 | 22–34 | 16–29 | 25–40 |
| Car wash | 60–100 | — | 50–100 | — | 90–100 | — | 30–45 |
Sources, all checked October 6, 2026: HCVT (table credited to Porto Leone Consulting, dated July 17, 2025); Miller Cooper; Engineered Tax Services; Overline; Modern CFO Cost Segregation Allocation Benchmarks, version 2026.1.0, maintained by Matthew Gigantelli, ASCSP M009-25 (CC BY 4.0); Cost Seg Smart; AE Tax Advisors. A dash means the source has no selected row for that type in this comparison. Where a source has several rows for one type, the table uses its low-rise office, its 5–20 or 6–20 unit multifamily, and its full-service restaurant and hotel rows. Miller Cooper gives one range for offices and retail. Engineered Tax Services gives one range for all residential property. HCVT groups car washes with RV parks and storage, and Engineered Tax Services groups them with gas stations. HCVT's broad "Residential Rental" row is mapped to single-family rental. Cost Seg Smart does not explicitly define the percentage denominator on its benchmark page; the other definitions are in the download.
Now read across any row. The labels are broad, and some sources group several kinds of property together.
| Building type | Sources | Lowest figure | Highest figure | Middle of the ranges | Do two ranges fail to overlap? |
|---|---|---|---|---|---|
| Warehouse | 6 | 10% | 30% | 13.5% to 28% | No |
| Office | 6 | 10% | 40% | 15.5% to 31% | Yes: 15–25 vs. 26–38 |
| Retail strip | 6 | 12% | 40% | 20% to 35.5% | No |
| Multifamily | 6 | 14% | 50% | 20% to 35% | Yes: 15–25 vs. 28–38 |
| Single-family rental | 5 | 9% | 40% | 20% to 34% | No |
| Hotel | 5 | 22% | 45% | 25% to 42% | No |
| Restaurant | 6 | 16% | 50% | 25% to 43% | Yes: 16–29 vs. 32–44 |
| Medical office | 5 | 16% | 40% | 22% to 38% | Yes: 16–29 vs. 30–38 |
| Car wash | 4 | 30% | 100% | 55% to 100% | Yes: 30–45 vs. 90–100 |
Source: CostSegregationMatch analysis of the seven sources in Table 3, checked October 6, 2026. "Middle of the ranges" is the median of the published low ends and the median of the published high ends.
In 5 of the 9 building types, two published ranges do not overlap at all. Take restaurants. One table tops out at 29%. Another starts at 32%. Different samples or methods can produce ranges that do not overlap. These ranges do not establish one measured national range.
Among the nine displayed building types, car washes have the widest span: one source says 30% to 45%, and another says 90% to 100%.
Why don't the tables agree?
Because they are not the same kind of evidence. Here is how each publisher describes its own numbers:
- Modern CFO offers the most open file: 45 building types, a version number and a free download. Its page says the bands reflect "expert judgment." A companion page by its maintainer says the file holds no study-by-study records, sample sizes or collection dates.
- Overline says its ranges come from its own database of 1,000+ studies. It does not publish the records.
- Cost Seg Smart says its ranges come from its own delivered studies where it has enough to measure, and from models otherwise.
- AE Tax Advisors says these are the ranges it sees in practice.
- HCVT credits its table to another consulting firm.
- Miller Cooper and Engineered Tax Services do not give the underlying sample or calculation for these ranges.
We found no public study-by-study data file behind the ranges from these seven sources. That is the gap the case reports in Table 1 start to fill. They are few, but you can check every dollar.
One more check. The Manning restaurant's 57.3% is higher than the top of every published restaurant range in our benchmark file. The highest of those is 50%.
Why do mobile home parks and farms look so different?
Four mobile home park case reports assigned 97.6% to 100.0% of depreciable basis to property with lives of 15 years or less. In those reports, 15-year property alone was 61.0% to 99.6% of the basis. Two farm reports also used 20-year property, so we show them on their own and keep all six out of the 31.3% median.
| Case report | Depreciable basis | Share in lives of 15 years or less | Share in lives of 20 years or less |
|---|---|---|---|
| Mobile home park, Okawville, IL | $388,000.00 | 100.0% | 100.0% |
| Mobile home park, Freeport, IL | $207,730.60 | 99.6% | 99.6% |
| Mobile home park, Freeport, IL | $1,811,378.70 | 97.6% | 97.6% |
| Mobile home park, Marion, IL | $1,800,000.00 | 98.9% | 98.9% |
| Farm and residential improvements, Talent, OR | $722,907.76 | 68.4% | 75.2% |
| Farm, Talent, OR | $3,570,454.91 | 41.5% | 94.9% |
Source: CostSegregationMatch analysis of Engineered Tax Services' published case reports. Each report name links to its source page. All pages checked October 6, 2026.
A 100.0% here does not mean the whole purchase price can be written off fast. It is 100.0% of the depreciable basis, and land is not in that number.
The two Freeport reports and the two Talent reports show different dollars, so we kept them as separate records. We could not confirm whether each pair covers the same property.
How much is cost segregation worth in dollars?
On $1 million of commercial depreciable basis, excluding land, with 30% reclassified, the modeled first-year depreciation deduction is $308,239 with a study and $11,770 without, for property placed in service in July under 100% bonus depreciation. That is a head start on deductions you would get anyway, not extra deductions. In the no-sale model, each $100,000 moved is worth about $19,510 at a 37% federal tax rate and a 5% discount rate, with every deduction fully usable.
Start with the first-year deduction, because that is the number used in the provider examples we reviewed.
| Share reclassified | Commercial building, with study | Commercial, no study | Residential rental, with study | Residential, no study |
|---|---|---|---|---|
| 20% | $209,416 | $11,770 | $213,336 | $16,670 |
| 30% | $308,239 | $11,770 | $311,669 | $16,670 |
| 31.3% (median of the 13 reports) | $321,463 | $11,770 | $324,828 | $16,670 |
| 40% | $407,062 | $11,770 | $410,002 | $16,670 |
Source: CostSegregationMatch calculation from IRS Publication 946 depreciation tables, October 2026. Commercial means a 39-year building. Residential rental means a 27.5-year building. The 13 case reports are all commercial; the residential column applies their median share only as an illustration. Calculations use the exact median, 31.3381867359502%, before rounding the dollar result.
Say you buy a commercial building with $1 million of depreciable basis and place it in service in July. With no study, your first-year deduction is $11,770. If a study reclassifies 31.3%, the sample median, it is $321,463.
The month matters more than you might think. A building placed in service in January gets $24,610 in year one without a study. In December it gets $1,070. That is why we show dollars and not "27 times more."
Is the first-year tax cut the real savings?
No. It is mostly timing.
A study does not create new deductions. It lets you take them sooner. You give up the small yearly deductions you would have taken on those dollars for the next 39 years.
So the fair question is: what is it worth to get the money early? That is present value. It shrinks future dollars back to what they are worth today. The discount rate is the yearly rate used for the shrinking. At 5%, $105 a year from now counts as $100 today.
| Discount rate | Commercial building (39-year) | Residential rental (27.5-year) |
|---|---|---|
| 3% | $14,625 | $11,360 |
| 5% | $19,510 | $15,874 |
| 7% | $22,441 | $18,923 |
| 10% | $24,835 | $21,774 |
| First-year tax cut, for comparison | $36,565 | $36,383 |
Source: CostSegregationMatch calculation from IRS Publication 946 depreciation tables, October 2026. Federal tax only. Property placed in service in July. Assumes the owner can use every deduction at the stated tax rate and keeps the property through its full recovery period. Tax savings occur at each year-end and are valued at the start of the first tax year.
At a 5% discount rate, the present value is $19,510 on a commercial building. The first-year tax cut is $36,565. So the lasting gain is about half of the first-year number: 53% for commercial property and 44% for residential.
Back to the $1 million building at the 31.3% median. The first-year federal tax cut at 37% is about $114,587. The present-value gain is about $61,141.
Has that number changed since the old rule of thumb?
Yes. In August 2004, the Journal of Accountancy printed this figure: about $16,000 of present-value savings for each $100,000 moved from 39-year to 5-year property, at a 35% tax rate and 5% discount rate. The article credited BKD LLP.
Using no bonus depreciation and tax savings at each year-end, our calculation is $15,369 for January placement and $15,733 for July. Those are our timing assumptions; the article did not state its timing. With 100% bonus and a 37% rate, our July no-sale model gives $19,510.
What if you sell?
A sale can reduce the benefit. Depreciation recapture can make some of the sale gain taxable as ordinary income.
Here is one illustration, using the same 37% tax rate and 5% discount rate. Assume the full $100,000 was assigned to personal property and written off in year one. At sale, all $100,000 is taxed as ordinary income at 37%; the slower building schedule's depreciation is taxed at an assumed 25%. A sale in December of the fifth tax year leaves a present-value gain of $4,879, 75% less than the no-sale model. A December sale in the tenth tax year leaves $9,417. Year one is the July placed-in-service year, and the model includes the mid-month deduction for the sale year.
The gain on each asset, its tax class, the tax rate and the sale date change the result. The actual rules are in IRS Publication 544. And if rental-loss limits stop you from using the deduction in year one, the gain shrinks too. Those rules are in IRS Publication 925.
What did the 2025 tax law change?
Qualified property acquired and placed in service after January 19, 2025 can get a 100% first-year allowance, known as bonus depreciation. For otherwise eligible property acquired after September 27, 2017 and before January 20, 2025, the old schedule generally gives 40% when placed in service in 2025 and 20% when placed in service in 2026. On the same modeled $1 million building, 100% versus 40% bonus means a first-year deduction of $308,239 versus $155,239.
This is why the topic is hot again. Bonus depreciation is what turns "5-year property" into "write it all off this year." From 2023 through early 2025, that bonus was shrinking. The July 2025 law brought back 100% and made it permanent. The IRS explains the rule in Publication 946 and in Notice 2026-11, released January 14, 2026.
| Bonus rule | Bonus rate | First-year deduction |
|---|---|---|
| Placed in service in 2022 | 100% | $308,239 |
| Placed in service in 2023 | 80% | $257,239 |
| Placed in service in 2024 | 60% | $206,239 |
| Placed in service in 2025, acquired before January 20, 2025 | 40% | $155,239 |
| Acquired and placed in service after January 19, 2025 | 100% | $308,239 |
| Placed in service in 2026, acquired before January 20, 2025 | 20% | $104,239 |
| No bonus depreciation | 0% | $53,239 |
Source: CostSegregationMatch calculation from IRS Publication 946 tables, October 2026. Bonus rates from IRS Notice 2026-11. The reclassified 30% is split two-thirds 5-year property and one-third 15-year property. The examples assume eligible assets, the stated bonus rate and the half-year rule for short-life property. Old-rule acquisitions are after September 27, 2017; the special long-production and aircraft rules are outside this comparison.
Where one owner on the old 40% rule deducts $1 in year one, an owner on the new rule deducts about $2.
The break is not small for the Treasury either. The Joint Committee on Taxation estimated that full expensing for certain business property would reduce federal revenue by about $363 billion over fiscal years 2025–2034, compared with prior law. That is a forecast for the broader tax rule, not measured cost segregation savings.
Ordering a study today does not reset a property's acquisition or placed-in-service date. Eligible property placed in service in 2021 could already qualify for 100% bonus under the older law. A catch-up study must use the rules that apply to the property's original dates.
How much does a cost segregation study cost?
CostSegregation.com displayed a $495 price for self-guided software on a rental with $750,000 of tax basis, excluding land. Engineered Tax Services says most professionally prepared studies run $3,000 to $12,000. In 2004, the Journal of Accountancy put a typical study at $10,000 to $25,000.
| Offer | Published price | What the price covers |
|---|---|---|
| CostSegregation.com software, residential | $495 | An example at $750,000 of tax basis, not counting land. You enter the property details yourself. |
| CostSegregation.com software, commercial | $1,295 | An example at $1 million of tax basis, not counting land. |
| R.E. Cost Seg Rapid Report | From $950 | Residential rentals under 3,500 square feet, with up to four substantially similar units. Main-page limits: $1.2 million of basis and $50,000 of renovations. No site inspection; see the boundary note below. |
| R.E. Cost Seg Fully Engineered Study, residential | From $2,320 | Virtual inspection included. An in-person visit costs extra. |
| R.E. Cost Seg Fully Engineered Study, commercial | From $2,730 | Virtual inspection included. An in-person visit costs extra. |
| Engineered Tax Services, stated range | $3,000 to $12,000 | The firm's own description of most professionally prepared studies. Not a quote. |
Source: each provider's own page, checked October 6, 2026. These are examples and starting prices, not quotes for your property.
R.E. Cost Seg lists optional Form 3115 preparation, for the IRS form used to change an accounting method, at $600 per study. The software has a $1.5 million basis limit, excluding land: its homepage says "up to," while its FAQ says "under." The Rapid Report page says $1.2 million of basis or less and $50,000 of renovations or less, while its FAQ says "under" both limits. Confirm eligibility with the provider if you are exactly at a limit. Check these terms before you compare prices.
To see what each of these offers includes, you can compare published study offers side by side.
Does a study pay for itself "10 times over"?
Sometimes, and sometimes not. It depends on the fee, the building, and whether you can use the deduction.
Take $1 million of commercial basis. If a study reclassifies 20% to 40%, the present-value gain is $39,020 to $78,040, using the $19,510 figure from Table 7. Set that against a fee of $3,000 to $12,000:
- Higher modeled ratio: $78,040 ÷ $3,000 = about 26 to 1.
- Lower modeled ratio: $39,020 ÷ $12,000 = about 3 to 1.
That model uses a 37% federal tax rate, a 5% discount rate, July placement and 100% bonus. It assumes every deduction is fully usable at that rate and no sale. The ratio is benefit divided by fee, before subtracting the fee or accounting for its tax treatment. A lower tax rate, a quick sale, or a deduction you cannot use yet all pull the ratio down.
If you are weighing a study for your own property, Find My Cost Seg Provider compares published study offers by price, property fit and workload. You do not need an account or contact details to use it.
Does the IRS accept cost segregation?
The IRS publishes a Cost Segregation Audit Techniques Guide, Publication 5653, to help its examiners review these studies. The guide lists 13 principal elements of a quality study. It also says it is not an official statement of the law, so treat it as the examiner's checklist, not a stamp of approval.
In plain terms: the IRS expects to see these studies, and it has written down what a careful one looks like. What it reviews is the support behind each number, such as cost records, site inspection and the reasoning for each asset's class. None of the pages that publish the 20% to 40% figure cite the IRS for it.
What do these numbers not tell you?
They describe 20 public case reports, a comparison of seven benchmark sources and a handful of price pages, all checked on October 6, 2026. They do not measure every study done in the United States. No public source we found counts how many studies are done, what the average fee paid is, or how often a study is audited.
Keep these limits next to any number you quote:
- One provider picked the cases. Engineered Tax Services chose which reports to publish.
- Thirteen is a small group. It is enough to show a range. It is not enough to rank building types or states.
- We did not see the studies. We read the published summaries. We did not see the engineering files or any tax return.
- A share is not a saving. A case-report percentage here is depreciable basis assigned to a faster class. It is not a refund.
- Benchmark tables are claims. They come from firms that sell studies or tax services.
- The dollar math is a model. It uses IRS tables and stated assumptions. Your month, tax rate, state taxes and plans will change it.
- The case reports cover nine states. That helps you find a nearby example. It does not make a state average.
This page is for learning, not tax advice. A tax professional decides what fits your return.
How did we build this?
We collected the 20 cost segregation case reports shown before "Load More" on one provider's public index, read each source page on October 6, 2026, and computed every share from the published dollars. Nineteen of the 20 gave enough figures to compute a share. This is the first yearly edition of our State of Cost Segregation report.
The case reports. Engineered Tax Services' case-study page showed 21 cards before its "Load More" button. One was about a research tax credit. The other 20 are in our file. They were published between January 27 and September 25, 2026. The provider's library holds older reports that we did not include.
The math. For each report, we added the dollars in each fast class and divided by the depreciable basis. We computed the median, mean and weighted share before rounding. Public figures show one decimal.
What we set aside, and why.
- The Red Oak, Texas apartment report lists $52,008,500 of depreciable basis but no dollars by class. It stays in the file with blank results.
- The Turlock pharmacy page shows two sets of percentages. Its summary reads 47%, 30% and 23%. Its detailed table reads 10.47%, 11.30% and 78.23%, and the dollars add up to the basis. We used the dollars, which give 21.8%.
- The Okawville park's web address says Elmhurst. The page itself says Okawville.
- In every usable report, the class amounts add up to the stated basis within 7 cents.
- We looked for a second provider with comparable pages. We did not add one because we could not confirm its "cost basis" leaves out land.
The benchmark comparison. We read nine publishers' pages. Our property-type comparison uses seven published guides; the download also records overall claims from KBKG and R.E. Cost Seg. It holds 106 benchmark records from 104 distinct source statements. Miller Cooper's office-and-retail range is mapped twice, as is Engineered Tax Services' residential range. We kept each publisher's definition of what the percent is "of," or marked it as unstated. Tables 3 and 4 show the nine mapped building types with at least four sources. The 12-row summary also lists self-storage, manufacturing and overall claims. The file is a selected compilation, not every range on every source page. The mapping and source-statement IDs are columns in the file, so you can check or change them.
The dollar math. We used the first-year percentages in IRS Publication 946: 20% for 5-year property and 5% for 15-year property under the half-year convention, and the mid-month tables for 27.5-year and 39-year buildings. These are timing rules for when depreciation starts and ends. The model uses the general system within MACRS, the IRS depreciation system, and a full calendar tax year. It assumes eligible property, the stated bonus rate and no Section 179 deduction, a separate way to deduct eligible business property. For bonus rates below 100%, we split the reclassified share two-thirds 5-year and one-third 15-year. In the 13 case reports the actual split was 56.8% and 43.2%. At 100% bonus the split does not change the answer.
The half-year rule must apply to the short-life assets. For the December examples, this requires enough other qualifying assets placed in service earlier in the year to keep the last-quarter share of relevant asset basis at 40% or less. Building basis is excluded from that test. If the mid-quarter rule applies, the first-year figures change.
The present-value model puts every tax saving at the end of its tax year and values it at the start of the first year. It assumes July placement, 100% bonus on the reclassified amount, full use of every deduction, the same federal marginal tax rate each year, and no sale or early disposal unless a sale is shown. A marginal rate is the tax rate on the next dollar of taxable income. The model leaves out state tax and other tax effects. The 5% discount rate is our assumption, not an IRS rate.
To reproduce a first-year figure, multiply the building basis by this sum: the reclassified share times the bonus rate; plus the unbonused reclassified share times the short-life first-year rate; plus the remaining building share times its first-year rate. At a 30% reclassified share and 40% bonus, this is $200,000 × (40% + 60% × 20%) + $100,000 × (40% + 60% × 5%) + $700,000 × 1.177% = $155,239.
To reproduce present value, multiply each year's tax saving by 1 ÷ (1 + the discount rate) raised to that year's number, then add those values. Subtract the present value of the building schedule from that of the full first-year write-off. The schedule download gives every annual IRS rate, including the final year. The findings file gives the sale-year adjustment and the January and July historical calculations. We use full-precision shares and round only the results.
To reproduce the 31.3%. Open the case-report file. Keep rows where status is included and group is commercial_building. Divide allocated_15yr_or_less_usd by depreciable_basis_usd and multiply by 100. Sort. Take the seventh.
More on how we work: our methodology and editorial standards. Found a figure that does not match its source? Send the record ID and the source through our contact page. We date and show corrections.
How do you cite this page?
Use the report name, the publisher and the date of the version you used. Keep the words "13 published commercial building reports" next to the 31.3% so its scope travels with it.
Ready-to-copy citation:
CostSegregationMatch. (2026). Cost Segregation Statistics (2026): The 20–40% Rule, Checked. State of Cost Segregation, 2026 edition. Data checked October 6, 2026. https://costsegregationmatch.com/research/cost-segregation-statistics/
Reuse: you may reuse CostSegregationMatch's own tables, calculations and charts with credit to CostSegregationMatch. Figures taken from other publishers stay theirs, and their terms apply. Rows from Modern CFO are used under CC BY 4.0 and need this credit: "Modern CFO Cost Segregation Allocation Benchmarks, version 2026.1.0, maintained by Matthew Gigantelli, ASCSP M009-25. Source: https://freecostseg.com/resources/cost-segregation-benchmarks/". We reproduced accelerated-range endpoints, mapped property labels and calculated comparisons. Keep that source credit, license link and change note with reused material.
Where can you download the data?
Eight CSV files hold everything on this page. They are free, with no form and no email. Every row carries its source and check date.
- cost-segregation-case-reports-2026.csv: the 20 case records
- cost-segregation-findings-2026.csv: each finding with its math
- cost-segregation-published-benchmarks-2026.csv: 106 benchmark records from 9 publishers, representing 104 distinct source statements
- cost-segregation-benchmark-summary-2026.csv: 12 summaries, with the nine types in Table 4 marked
- first-year-deduction-per-1m-2026.csv: first-year deductions behind Tables 6 and 8
- present-value-per-100k-reclassified-2026.csv: present values behind Table 7
- cost-segregation-published-fees-2026.csv: published prices behind Table 9
- cost-segregation-model-schedules-2026.csv: all 228 annual-rate rows used in the IRS-table models
Cost segregation statistics FAQ
Is it worth it to do cost segregation?
It depends on your building, your tax rate and your plans. On $1 million of commercial basis with 20% to 40% reclassified, the modeled present-value gain is about $39,020 to $78,040 at a 37% federal rate and 5% discount rate, before the study fee. That assumes July placement, 100% bonus, full use of the deductions and no sale. If you cannot use the deduction soon, or you plan to sell within a few years, the gain is smaller. Ask your tax preparer to run it for your return.
What is the average cost segregation percentage?
We found no public national average. In 13 published commercial building reports from Engineered Tax Services, the median was 31.3% of depreciable basis and the mean was 34.0%. In the nine-type comparison, published "typical" ranges from seven sources sit between 9% and 50% for the eight types other than car washes, whose ranges go as high as 100%.
Is the 20% to 40% rule an IRS rule?
No source we checked ties it to the IRS. It appears on provider pages as an average, without a cited source. In the 13 case reports we analyzed, nine fell inside 20% to 40% and four were above it.
What are the rules for cost segregation bonus depreciation in 2026?
Qualified property acquired and placed in service after January 19, 2025 can get 100% bonus depreciation, according to IRS Publication 946 and IRS Notice 2026-11. Eligible property acquired after September 27, 2017 and before January 20, 2025 generally gets 20% if first placed in service in 2026; special long-production and aircraft rules differ. The study date does not set the bonus rate. A study finds which parts of a building qualify. It does not make the whole building eligible.
Can I do cost segregation on a building I bought four years ago?
Often, yes. A study can identify depreciation missed in earlier years, and correcting an established depreciation method generally uses IRS Form 3115. The original acquisition and placed-in-service dates still control the bonus rules: eligible property placed in service in 2022 could qualify for 100% under the older law. Ask your tax preparer what filing your case needs.
Can you explain cost segregation in a simple way?
A building is not one thing. It is a shell plus carpet, cabinets, wiring for equipment, a parking lot and more. Tax rules let some of those parts be written off in 5, 7 or 15 years, where the shell takes 27.5 or 39. A cost segregation study sorts the parts into the right piles. In the 13 published commercial building reports we reviewed, the fast piles held a median of 31.3% of depreciable basis, excluding land.
How much does a cost segregation study cost?
CostSegregation.com displayed a $495 price for self-guided software on a rental with $750,000 of tax basis, excluding land. Reports with engineering review start at $950 to $2,730 at one provider, and one national firm says most professionally prepared studies run $3,000 to $12,000. Tax-form help can cost extra.
Is cost segregation a tax deduction or a tax deferral?
It speeds up depreciation deductions. In our no-sale model, total deductions stay the same, but more come early. At a 37% federal rate and 5% discount rate, moving $100,000 of a July-placed commercial building to 100% bonus gives a $19,510 present-value gain, compared with a $36,565 first-year tax cut. The model assumes every deduction is fully usable at that rate.
How many cost segregation studies are done each year?
We found no public count. This page covers 20 published case reports and compares ranges from seven benchmark sources. It does not estimate how many owners use a study or how large the market is.
Sources
All sources were checked on October 6, 2026.
- Engineered Tax Services, case-study index. https://engineeredtaxservices.com/case-studies/
- Engineered Tax Services, case report ETS01: Fast-food restaurant, Manning, SC (published 2026-09-25). https://engineeredtaxservices.com/case-studies/cost-segregation-for-a-fast-food-restaurant-in-manning-south-carolina/
- Engineered Tax Services, case report ETS02: Retail pharmacy, Turlock, CA (published 2026-08-17). https://engineeredtaxservices.com/case-studies/case-study-cost-segregation-for-retail-pharmacy-property-in-turlock-california/
- Engineered Tax Services, case report ETS03: Multifamily apartments, Red Oak, TX (published 2026-07-08). https://engineeredtaxservices.com/case-studies/cost-segregation-study-for-a-luxury-apartment-community-in-red-oak-texas/
- Engineered Tax Services, case report ETS04: Warehouse / manufacturing, Corona, CA (published 2026-03-24). https://engineeredtaxservices.com/case-studies/case-study-cost-segregation-amazing-warehouse-manufacturing-property-corona-california/
- Engineered Tax Services, case report ETS05: Preschool, Wesley Chapel, FL (published 2026-03-04). https://engineeredtaxservices.com/case-studies/case-study-cost-segregation-pre-school-property-wesley-chapel-florida/
- Engineered Tax Services, case report ETS06: Preschool, Cape Coral, FL (published 2026-03-04). https://engineeredtaxservices.com/case-studies/case-study-cost-segregation-analysis-of-a-pre-school-in-cape-coral-florida/
- Engineered Tax Services, case report ETS07: Mobile home park, Okawville, IL (published 2026-02-12). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-mobile-home-park-elmhurst-il/
- Engineered Tax Services, case report ETS08: Medical office, Floyds Knobs, IN (published 2026-02-08). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-medical-office-floyds-knobs-in/
- Engineered Tax Services, case report ETS09: Gym, Joplin, MO (published 2026-02-04). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-gym-joplin-mo/
- Engineered Tax Services, case report ETS10: Warehouse / office, Joliet, IL (published 2026-02-01). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-warehouse-office-joliet-illinois/
- Engineered Tax Services, case report ETS11: Mobile home park, Freeport, IL (published 2026-01-28). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-mobile-home-park-freeport-il/
- Engineered Tax Services, case report ETS12: Warehouse / office, Elmhurst, IL (published 2026-01-27). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-warehouse-office-elmhurst-il/
- Engineered Tax Services, case report ETS13: Warehouse / storage, Katy, TX (published 2026-01-27). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-warehouse-storage-katy-tx/
- Engineered Tax Services, case report ETS14: Farm and residential improvements, Talent, OR (published 2026-01-27). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-farm-property-talent-or/
- Engineered Tax Services, case report ETS15: Office, Tampa, FL (published 2026-01-27). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-office-property-tampa-fl/
- Engineered Tax Services, case report ETS16: Mobile home park, Freeport, IL (published 2026-01-27). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-mobile-home-park-freeport-illinois/
- Engineered Tax Services, case report ETS17: Mobile home park, Marion, IL (published 2026-01-27). https://engineeredtaxservices.com/case-studies/case-study-cost-segregation-analysis-mobile-home-park-marion-il/
- Engineered Tax Services, case report ETS18: Farm, Talent, OR (published 2026-01-27). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-farm-talent-or/
- Engineered Tax Services, case report ETS19: Dental office, Belmont, NH (published 2026-01-27). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-dental-office-belmont-nh/
- Engineered Tax Services, case report ETS20: Auto service facility, Richardson, TX (published 2026-01-27). https://engineeredtaxservices.com/case-studies/cost-segregation-analysis-auto-service-facility-richardson-tx/
- KBKG, "Cost Segregation" (page modified May 1, 2026). https://www.kbkg.com/costsegregation
- Miller Cooper, "Cost Segregation." https://millercooper.com/cost-segregation/
- HCVT, "Viability of Cost Segregation Studies for Tax Year 2025" (July 17, 2025). https://www.hcvt.com/alertarticle-Cost-Segregation
- Engineered Tax Services, "Cost Segregation Services" (page modified September 1, 2026). https://engineeredtaxservices.com/services/cost-segregation/
- Overline, "How Much Does Cost Segregation Actually Save?" (January 12, 2026). https://overlineiq.com/blog/how-much-does-cost-segregation-save
- Modern CFO Cost Segregation Allocation Benchmarks, version 2026.1.0, maintained by Matthew Gigantelli, ASCSP M009-25 (CC BY 4.0). https://freecostseg.com/resources/cost-segregation-benchmarks/
- Overline, "Cost Segregation Allocation Benchmarks by Property Type" (March 18, 2026). https://overlineiq.com/blog/cost-segregation-benchmarks-8000-studies
- Cost Seg Smart, "Cost Segregation Benchmarks." https://costsegsmart.com/cost-segregation-benchmarks/
- AE Tax Advisors, "Cost Segregation ROI by Property Type" (updated August 17, 2026). https://www.aetaxadvisors.com/cost-segregation-roi-by-property-type/
- R.E. Cost Seg, "Cost Segregation Reclassification: The 25–45% Rule of Thumb," Mitchell Baldridge (September 11, 2026; article body). https://www.recostseg.com/post/reclassification-percentage
- Internal Revenue Service, Publication 946 (2025), How To Depreciate Property. https://www.irs.gov/publications/p946
- Internal Revenue Service, Notice 2026-11 (released January 14, 2026). https://www.irs.gov/pub/irs-drop/n-26-11.pdf
- Internal Revenue Service, Publication 5653 (2-2025), Cost Segregation Audit Techniques Guide; February 6, 2025 cover date, page 35 and retail asset matrix. https://www.irs.gov/pub/irs-pdf/p5653.pdf
- Internal Revenue Service, Publication 925, Passive Activity and At-Risk Rules. https://www.irs.gov/publications/p925
- Internal Revenue Service, Publication 544, Sales and Other Dispositions of Assets. https://www.irs.gov/publications/p544
- Jay A. Soled and Charles E. Falk, "Cost Segregation Applied," Journal of Accountancy (August 2004). https://www.journalofaccountancy.com/issues/2004/aug/costsegregationapplied/
- Joint Committee on Taxation, JCX-35-25 (July 1, 2025), page 2, Chapter 3, Subchapter A, item 1, "Full expensing for certain business property." Estimate for Senate-passed legislation, fiscal years 2025–2034, against the present-law baseline: −$362,650 million. https://www.jct.gov/getattachment/eb21dc77-6439-4fc3-8f5d-fc23a8c377e0/x-35-25.pdf
- CostSegregation.com, from KBKG. https://www.costsegregation.com/
- R.E. Cost Seg, Rapid Report. https://www.recostseg.com/services/rapid-report
- R.E. Cost Seg, Fully Engineered Study. https://www.recostseg.com/services/fully-engineered-study
- R.E. Cost Seg, services overview. https://www.recostseg.com/services
- CostSegregation.com, FAQ; software basis limit. https://www.costsegregation.com/faq/
- Internal Revenue Service, Revenue Procedure 2025-32, section 4.01; 2026 individual marginal rates. https://www.irs.gov/pub/irs-drop/rp-25-32.pdf
- Internal Revenue Service, Instructions for Form 3115, Application for Change in Accounting Method. https://www.irs.gov/instructions/i3115
- Modern CFO, Cost Segregation Allocation Benchmarks methodology, version 2026.1.0. https://freecostseg.com/resources/cost-segregation-benchmarks/methodology/
- Creative Commons, Attribution 4.0 International license. https://creativecommons.org/licenses/by/4.0/