How the R&D Credit Is Actually Calculated: Two Methods, One Election, and Why Year One Looks Different

TL;DR

• The federal R&D credit is calculated one of two ways: the regular credit (20% of qualified spending above a historical base) or the alternative simplified credit (14% of qualified spending above half of your three-year average). Both are incremental, so the credit rewards spending above a baseline, not total spending.
• A company with no qualified research expenses in one of the three prior years uses a 6% rate under the ASC, which is why a first-year claim often looks smaller than the same activity will the following year.
• Because the credit and the Section 174A deduction interact, every filer either reduces the deduction by the credit amount or makes the Section 280C election for a reduced credit. Which is better depends on the company's tax rate and situation.

The number on the return is not "20% of what we spent"

Ask most owners how the R&D credit works and they will say something like "you get back a percentage of what you spend on research." That is close enough to be misleading. The federal credit for increasing research activities under IRC Section 41 is, as the name says, a credit for increasing research. It is computed on qualified research expenses (QREs) above a base amount, and the base amount depends on the company's own history.

That design has two practical consequences. First, two companies with identical research spending this year can end up with different credits. Second, the first year a company claims can look modest compared with the second and third, even when the underlying engineering work has not changed.

The activities that generate QREs still have to pass the qualification rules in Treasury Regulation 1.41-4. This post assumes that work is done and focuses on what happens to the dollars afterward.

Method one: the regular credit

Section A of Form 6765 computes the regular credit. In broad strokes:

The credit is 20% of the current year's QREs that exceed a base amount. The base amount is the company's fixed-base percentage multiplied by its average annual gross receipts for the four preceding tax years. The fixed-base percentage is, for a company that did not exist in the mid-1980s, set at 3% for its first five years with QREs and then recalculated from the company's own research-to-revenue ratio in later years, capped at 16%. And regardless of how the math comes out, the base amount can never be less than 50% of the current year's QREs, so the regular credit effectively tops out at 10% of total QREs.

The regular credit can produce a larger number than the ASC for some companies, particularly those whose research spending is high relative to revenue. It also requires gross receipts history and, for mature companies, a fixed-base computation that reaches back decades. For many companies in the $3M to $15M range, the data burden is the reason it is rarely the first choice.

Method two: the alternative simplified credit

Section B of Form 6765 computes the alternative simplified credit (ASC), added to the statute in Section 41(c)(4). The formula is easier to hold in your head:

The credit is 14% of the current year's QREs that exceed 50% of the average QREs for the three preceding tax years. Gross receipts do not enter into it.

If the company had no QREs in any one of those three preceding years, the rate drops to 6% of current-year QREs, with no base subtraction. This is the rule behind the "year one looks different" pattern. A company that has done qualifying work for years but never identified or tracked it may still have prior-year QREs that can be reconstructed and used in the base, which is one of the reasons a well-run study looks at prior years rather than only the current one. Whether that reconstruction is supportable depends on the records that exist.

Two other ASC rules matter. The election is made by completing Section B on a timely filed original return (including extensions), and it can only be made on an amended return if the company has not previously claimed the credit for that year by another method. Once made, the election carries forward to later years unless revoked. The IRS instructions themselves suggest computing both methods where eligible to see which produces the larger allowable credit.

A worked example (illustrative only)

Suppose a company has $400,000 of QREs this year and had $300,000, $320,000, and $340,000 in the three prior years. The three-year average is $320,000, and half of that is $160,000. Under the ASC, the credit would be 14% of ($400,000 minus $160,000), or $33,600.

Now suppose the same company had no QREs in one of those three prior years. The ASC rate becomes 6% of $400,000, or $24,000. Same work, same spending this year, different credit.

Under the regular credit, if the company is in its early years with a 3% fixed-base percentage and average gross receipts of $8 million, the base amount would be $240,000. That is above the 50%-of-QREs floor of $200,000, so it stands, and the credit would be 20% of ($400,000 minus $240,000), or $32,000.

These figures are hypothetical and simplified. They ignore the base-year consistency rule, controlled-group aggregation, state credits, and the Section 280C election described next. They are here to show the shape of the math, not to predict any company's result.

The Section 280C election: why the credit and the deduction are linked

Since the restoration of current-year expensing for domestic research under Section 174A, the interaction between the credit and the deduction has become relevant again for most companies. The rule in Section 280C(c) prevents a company from getting a full deduction and a full credit on the same dollars.

Per the current Form 6765 instructions, a company that does not make the election must reduce its domestic research or experimental expenditures deducted under Section 174A by the amount of the research credit. Alternatively, the company can check "Yes" on Item A of Form 6765 and elect a reduced credit: 15.8% instead of 20% under the regular method, or 79% of the computed ASC under the simplified method. The election has to be made on the original, timely filed return for the year, including extensions. It cannot be made or changed on an amended return, and once made it is irrevocable for that year.

Which path produces the better after-tax result depends on the company's marginal rate, its entity type, whether it has taxable income this year, and how the credit and deduction flow through to the owners. For pass-through entities in particular, the answer is not always obvious. This is a conversation to have with your CPA, since the choice lives on the return the CPA files.

The consistency rule most people have never heard of

Section 41(c)(5) requires that QREs in the base years be determined on the same basis as QREs in the credit year. In plain terms, if a company counts a category of spending as qualified this year, it generally has to count that same category in the prior years used to build the base, even for years that are otherwise closed. This is one of the quieter reasons that method choice and documentation matter: a base built inconsistently can overstate the credit, and a base that is reconstructed carefully tends to hold up.

What this means for an owner or CFO

None of this changes whether the work qualifies. It changes what the work is worth on the return, and it means the answer to "how much is this credit?" is not available from a single percentage. It depends on the company's own prior years, which method is elected, and how the credit interacts with the deduction.

Strata's initial assessment is $0, and it typically starts by looking at exactly these inputs: the current year, the prior years that can support a base, and which computation and election fit the company's situation. A full study typically runs four to eight weeks depending on how quickly records come together, with most of the work on Strata's side. Timeframes are typical averages and the engagement agreement governs.

If you want to understand which method and which election make sense for your company, start with a conversation.

Sources

• IRC Section 41, Credit for increasing research activities, including 41(c)(2) minimum base amount, 41(c)(3) fixed-base percentage, 41(c)(4) alternative simplified credit, and 41(c)(5) consistent treatment
• Instructions for Form 6765 (Rev. December 2025), IRS.gov, Item A (Section 280C election), Section A (regular credit), Section B (ASC)
• Treasury Regulation 1.41-4, Qualified research
• Treasury Regulation 1.41-9, Alternative simplified credit

This post is for informational purposes only and does not constitute tax or legal advice. Consult a qualified tax professional regarding your specific circumstances.