What are the Section 174 transition options under the OBBBA?

Two tracks. Small businesses — average annual gross receipts of $31 million or less — may apply Section 174A retroactively to tax years beginning after 2021 by amending 2022–2024 returns. All taxpayers, regardless of size, may instead deduct their remaining unamortized domestic Section 174 balances over one or two tax years beginning with the first tax year after 2024.

Who qualifies as a small business for retroactive Section 174A relief?

Taxpayers meeting the Section 448(c) gross-receipts test at a $31 million threshold — average annual gross receipts for the three prior tax years of $31 million or less, with aggregation of related entities under Sections 52 and 414. Tax shelters are excluded.

Do foreign research costs get the catch-up deduction?

No. The one- or two-year catch-up applies only to unamortized domestic Section 174 amounts from 2022–2024. Foreign SRE expenditures continue amortizing over their original 15-year schedules; the OBBBA left the foreign regime fully intact.

Is a Form 3115 required for the Section 174 catch-up?

The catch-up is implemented as an accounting-method change with a Section 481(a) adjustment, generally under automatic-consent procedures. The IRS's transition guidance provides streamlined mechanics — including statement-based elections in some cases — but the framework is method-change law, so timing and consistency rules apply.

Should a small business amend 2022–2024 or take the catch-up?

It depends on rates and posture. Amending recovers cash from closed years with overpayment interest and restores payroll-offset-friendly credit positions, but costs three amended filings and reopens those years to exam. The catch-up is one filing at current rates. A profitable small business that paid tax in 2022–2024 usually favors amending; loss companies usually don't.

Section 174 transition rules: amend, catch up in one year, or two

The restoration of research expensing answered the easy question — how are 2025 costs treated — and left the hard one: what happens to the roughly three years of domestic research costs capitalized under Section 174 during 2022–2024 and still sitting, mostly unamortized, on tax balance sheets. The OBBBA's answer is a menu. Small businesses (average annual gross receipts of $31 million or less) may apply new Section 174A retroactively to tax years beginning after 2021, amending their 2022–2024 returns. Every taxpayer, whatever its size, may instead deduct the remaining unamortized domestic balance over one or two tax years beginning with the first tax year after 2024.

Which option wins is a modeling exercise, not a rule of thumb. This guide walks the mechanics, works the numbers, and flags where each path fails.

The stranded-balance problem

A calendar-year taxpayer that spent evenly on domestic research during 2022–2024 recovered only a fraction through amortization. Under the five-year schedule with a half-year in year one, cumulative deductions by the end of 2024 were 50% of 2022 costs, 30% of 2023 costs, and 10% of 2024 costs. A company spending $1 million per year has $2.1 million of unamortized domestic balance entering 2025 — 70% of everything it spent. The transition rules exist to release that balance faster than the original schedules would (which run into 2029).

Foreign balances are excluded throughout. Costs on the 15-year foreign schedule keep amortizing as before; see the foreign research brief.

Option 1: small-business retroactive amendment

Taxpayers meeting the Section 448(c) gross-receipts test at a $31 million threshold — average annual gross receipts for the three preceding tax years, aggregated across commonly controlled entities — may elect to apply Section 174A to tax years beginning after December 31, 2021. In practice: amend 2022, 2023, and 2024 to deduct domestic SRE expenditures in the years incurred, unwinding capitalization entirely.

What the election buys:

Refunds with interest. Tax overpaid in 2022–2024 comes back with statutory overpayment interest — meaningful after several years at elevated rates.
Restored NOLs. A company that showed artificial taxable income in 2022–2024 because of capitalization may instead show losses, generating NOL carryforwards into 2025.
Clean 280C reconciliation. Credit computations for those years can be conformed to a deduction posture.

What it costs: three amended federal returns (plus state amendments in conforming states), flow-through complexity — partnerships and S corporations amend at the entity level and push adjustments to owners, who then amend personal returns — and re-opened examination exposure for each amended year. The decision framework is treated in depth in amended return vs. catch-up.

Option 2: the catch-up deduction — one year or two

All taxpayers may deduct the remaining unamortized domestic Section 174 balance over one or two tax years, beginning with the first tax year after December 31, 2024 — 2025 for calendar-year filers. One year means the entire balance deducts in 2025. Two years means it splits between 2025 and 2026 (ratably, under the transition guidance).

The catch-up is implemented through accounting-method-change mechanics: the taxpayer changes from the capitalization method to the 174A method, and the released balance is in substance a favorable Section 481(a) adjustment taken over one or two years. Automatic consent procedures apply; the filing mechanics are covered in Section 481(a) and Section 174 and the general Form 3115 guide.

Why take two years instead of one? Because a deduction is worth its marginal rate, and marginal rates are not constant:

A one-year catch-up that drives 2025 into a large loss converts the excess into an NOL usable only against 80% of future income.
A taxpayer expecting higher income (or higher rates) in 2026 gets more per dollar by deferring half.
Spreading can preserve the ability to absorb other 2025 attributes — expiring credits, foreign tax credits — that a mega-deduction would waste.

Worked example: three paths compared

Take a calendar-year C corporation with $1.5 million of domestic SRE spend in each of 2022, 2023, and 2024 ($4.5 million total). Amortization deducted $1.35 million through 2024, leaving a $3.15 million unamortized balance. Assume it paid 21% federal tax on the income inflated by capitalization and expects moderate 2025–2026 profits.

Total federal tax benefit is identical across options; timing, interest, and rate exposure differ:

Factor	Amend 2022–2024	One-year catch-up (2025)	Two-year catch-up (2025–26)
Deduction released	$3.15M into prior years	$3.15M in 2025	$1.575M each year
Cash timing	Refund claims, ~6–12 mo. processing	2025 return / reduced estimates	Spread over two filing cycles
Overpayment interest	Yes, on 2022–2024 refunds	No	No
Filings	3+ amended returns (plus states, owners)	One method change	One method change
Exam exposure	Reopens three years	Prospective only	Prospective only
NOL risk	Restores prior-year losses cleanly	May trap value in 80%-limited NOL	Mitigated by spreading
Available to	≤ $31M average gross receipts	All taxpayers	All taxpayers

Year deductions land under each transition option ($3.15M stranded balance)$

Illustrative: $1.5M annual domestic SRE spend 2022–2024; amendment column shows deductions relocated to prior years in aggregate.

Modeling: what actually decides it

Marginal rate in each landing year. For a C corporation the federal rate is flat, so rate arbitrage is mostly about usability — whether income exists to absorb the deduction. For pass-through owners, the analysis runs at individual rates, where the difference between a 37%-bracket year and a lower-income year is real money, and where Section 461(l) excess business loss limits can strand a one-year mega-deduction at the owner level.

NOL posture. If 2022–2024 tax was actually paid, amendment (if eligible) recovers hard dollars plus interest — hard to beat. If the company was in losses throughout despite capitalization, amendment merely reshapes NOL vintages, and the cheaper single method-change filing wins on administration. A one-year catch-up into a loss year is usually the worst configuration: it maximizes the 80%-limited NOL.

Refund interest vs. time value. Overpayment interest on 2022 tax refunded in 2026 accrues at rates that have run 7–8% for much of the period — often exceeding a corporate discount rate. This quietly tilts profitable small businesses toward amending.

Statute of limitations. The 2022 assessment/refund statute is live but aging; calendar-2022 returns filed in 2023 generally close in 2026. Amendment is a now-or-never decision for the earliest year.

Administrative cost. Three amended federal returns, state conformity analysis in every filing state, and — for flow-throughs — owner-level amendments can cost tens of thousands in professional fees. Against a modest refund, the catch-up's single filing can win on net.

When each option fails

Amendment fails when the taxpayer flunks the aggregated $31 million test, when 2022–2024 were loss years anyway, when owner-level amendment cascades are impractical, or when reopening years with other soft positions (worker classification, ERC claims) invites exam risk that outweighs the refund.
One-year catch-up fails when it creates an 80%-limited NOL, triggers 461(l) at the owner level, or wastes expiring attributes.
Two-year catch-up fails when income is high now and uncertain later — deferring half a deduction into a year that turns out to be a loss year forfeits time value for nothing.

The bottom line

The transition rules convert a stranded 70%-of-three-years balance into a scheduling decision. Profitable small businesses under the $31 million ceiling generally do best amending — refunds, interest, restored attributes — provided the administrative cascade is tolerable. Everyone else chooses between one year and two, and that choice is a straightforward present-value problem complicated by NOL limitations and owner-level loss caps. The only clearly wrong answer is not modeling it: the same $3.15 million deduction can be worth materially different amounts depending on where it lands.