Some of your design hours are research. Most of them are not.

Design that resolves a genuine technical uncertainty can meet the four-part test. Drafting, permitting, and code compliance cannot. Which hours fall on which side, and what your client agreements say, is the whole study.

Get your free credit estimate Run the estimator yourself

Photo by Marsumilae on Unsplash

Why some design work qualifies, and most does not

The credit does not reach design because it is difficult, or because it is billed. It reaches work on a business component where the information available to you did not establish whether the thing could be done, how to do it, or what the right design was - and where you resolved that by evaluating alternatives rather than by applying a method you already had.

That framing is why a transfer structure with no precedent in your details library, a facade assembly outside the tested range, or a strengthening scheme for a building whose real capacity had to be established by investigation are all candidates, and why a permit set, a code check against published tables, and a bid package are not. Each candidate is tested against the IRC §41 four-part test on its own facts.

Then there is the second half of the problem, which is contractual rather than technical. Almost everything a design practice does is paid for by a client, so §41(d)(4)(H) has to be answered agreement by agreement before anything enters a base. The reported outcomes for design firms cut genuinely both ways, which is why we read the documents rather than the letterhead.

Seven things that happened on a project last year

Not activity categories - situations. If any of these read like a job you actually ran, there is very likely a qualified business component underneath it.

“There was no detail in our library for that span.”

A long-span or transfer condition fell outside the framing you had built before. Alternative schemes were analysed and modelled, and one was carried only after the others were shown not to work.

Why it can qualify: Design uncertainty about the appropriate structural system, resolved by evaluating alternatives rather than by selecting a known detail.
“The building would not meet the energy target as drawn.”

Envelope build-ups, glazing ratios, and system alternatives were modelled and re-modelled against a calibrated simulation until the predicted performance held.

Why it can qualify: Improving performance is a permitted purpose, and iterative simulation against measured assumptions is a process of experimentation.
“The soil report changed the problem.”

Settlement and capacity uncertainty forced two foundation systems to be developed and compared before either could be priced, and one was abandoned on the analysis.

Why it can qualify: Capability and method uncertainty in the engineering sciences. The abandoned scheme is part of the process, not wasted effort.
“The existing structure was not what the drawings said.”

Capacity had to be established by investigation and analysis before a strengthening scheme could be selected, and the first scheme did not survive the second round of numbers.

Why it can qualify: The information available did not establish the method. Establishing it by test and analysis is what the statute asks for.
“The system had never been used in a building like this.”

An unconventional conditioning or thermal-storage approach was evaluated because its behaviour in this occupancy and climate could not be predicted from the published data.

Why it can qualify: Technological in nature and genuinely uncertain. Routing a conventional system through a tight ceiling is coordination, not research.
“We had to prove the assembly would behave.”

A connection or facade assembly outside the tested range was analysed, mocked up, and instrumented because the manufacturer’s data did not cover the condition.

Why it can qualify: Design uncertainty resolved by physical testing. The mock-up materials consumed are supplies.
“Standard isolation would not hold the vibration criterion.”

Sensitive-equipment criteria drove alternative isolation and stiffening schemes, modelled and then measured on site before the design was fixed.

Why it can qualify: A performance requirement the available information could not meet, resolved through a documented sequence of alternatives.

Illustrative situations, not client work. Whether any of them qualifies for you depends on your facts, your agreements, and your evidence.

A pencil and ruler resting on a dimensioned technical drawing — A dimensioned drawing under revision. Illustrative.Photo by Sven Mieke on Unsplash

The record is in the calculation set, not the time sheet

A design practice already produces the evidence a claim needs, in a form nobody files for tax: option studies with the schemes that were rejected, calculation sets with superseded revisions, model runs and their assumptions, mock-up photographs and instrumented results, and the meeting minute where the answer changed.

What is almost never in the record is the allocation - which hours went to the uncertain part of the project, and which went to documenting a settled design. A flat percentage applied across every project cannot answer that question, and it is the first thing an examiner asks about.

How substantiation is assembled

The design work that commonly qualifies

Representative activities we see meet the four-part test across structural, civil, building-systems, and building-science practice.

Structural systems & analysis

Developing framing, transfer, and lateral systems where the appropriate approach is genuinely uncertain and has to be established by analysis and comparison.

Geotechnical & foundation engineering

Foundation, retention, and ground-improvement schemes developed against settlement, capacity, and groundwater uncertainty.

Building envelope & building science

Thermal, moisture, air-barrier, and durability performance of assemblies that fall outside tested and published ranges.

Energy modelling & performance design

Simulation-driven design where the model is calibrated, alternatives are tested against it, and the design changes on the result.

Building systems development

Mechanical, electrical, plumbing, and controls approaches whose behaviour in the specific building cannot be predicted from manufacturer data.

Seismic & existing-structure assessment

Establishing the capacity of an existing structure by investigation and analysis, then developing and comparing strengthening schemes.

Civil, water & infrastructure engineering

Hydraulic, stormwater, treatment, and transport engineering where site conditions put the problem outside standard practice.

Acoustics, vibration & specialised criteria

Meeting performance criteria that standard details do not reach, resolved by modelling, mock-up, and measurement.

Design-phase mock-ups & testing

Physical and instrumented prototypes built during design to answer a question the drawings could not answer.

Typical QRE categories for a design practice

What spending counts toward the credit - tailored to a business whose costs are almost entirely people.

Typical QRE categories and their statutory basis
Expense category	What goes into the base
Technical wages§41(b)(2)(A)-(B)	W-2 wages for engineers, architects, designers, and analysts for time spent on qualified design work, plus direct supervision and direct support of it.
Contract research (65%)§41(b)(3)	65% of amounts paid to U.S. specialist consultants and testing laboratories for qualified research performed on your behalf under an agreement entered into before the work.
Supplies§41(b)(2)(C)	Materials consumed in mock-ups and test assemblies built during design. Usually a small line for a design office, and never depreciable equipment.
Computer rental§41(b)(2)(A)(iii)	Amounts paid for the right to use computers in qualified research, such as rented compute for a large simulation run. Per-seat software licences are not a §41 expense category.

General and illustrative. Only qualified research performed in the United States, Puerto Rico, or a U.S. possession is eligible, and contract research enters the base at 65% of the amount paid under §41(b)(3).

What the base usually looks like

Illustrative

A directional shape for a design practice, not a benchmark. Wages carry essentially the whole claim, which is precisely why the hour-level allocation has to be defensible rather than assumed.

Technical wages - Engineer, architect, and analyst time on projects with a real unknown.: 88%
U.S. contract research - Specialist consultants and test labs, in the base at 65% of amounts paid.: 9%
Supplies - Mock-up and test-assembly materials consumed during design.: 2%
Computer rental - Rarely material, and software seats do not count.: 1%

Where the line sits

Seat licences for CAD, BIM, and analysis software are not a §41 expense category at all, however large the invoice. Neither is rent, insurance, or the overhead loaded into your billing rate: a fully-loaded cost model will overstate the credit every time.

Your project accounting is usually good enough to support an hour-level allocation, because it already separates phases and tasks. The work is mapping those tasks to business components and to the point at which the uncertainty was actually resolved.

Full QRE rules, category by category

Exclusions to watch

Where design-firm claims actually fail

Two of these are about the contract and three are about the hours. A claim that ignores either half is the kind that does not survive an examination.

§41(d)(4)(H)

Work the client funded

Research is funded, and excluded, to the extent another party pays for it and you neither bear the financial risk of failure nor retain substantial rights in the results. For a design practice this is the whole ballgame, because almost everything is done under a client agreement. A fixed-fee agreement is a better starting position than a cost-reimbursement one, but it is only a starting position.

§1.41-4A(d)

A fixed fee is not by itself enough

The Eighth Circuit affirmed the Tax Court against a structural engineering firm in 2024 on contracts that were fixed-price, distinguishing proper performance from successful performance: an obligation to comply with applicable codes, or to work to a general standard of care, does not mandate success. What courts have looked for is express machinery - objective acceptance criteria, a right to reject non-conforming work, payment limited to accepted work, or a refund if benchmarks are missed.

§41(d)(3)(B)

Style, taste, and appearance

Design factors relating to style, taste, cosmetic appearance, or season are outside the permitted purposes. Massing studies, material palettes, and aesthetic iteration are the practice of architecture, not qualified research. Where an appearance decision created a real structural, envelope, or acoustic problem, it is the engineering that resolved it - not the palette - that is the candidate.

§41(d)(4)(A), (B), (D)

Documents, permitting, and code compliance

Producing construction documents for a design that is already settled, permit sets, code checks, routine calculation to published tables, bid support, and construction administration are not qualified research. Adapting an existing design to a particular client’s requirement is separately excluded. On most projects this is the majority of the fee, which is why an honest claim here is a fraction of the hours rather than a percentage of the practice.

§41(d)(4)(C), (F)

Repeat prototypes and offshore production

Rolling a proven building prototype across new sites is duplication of an existing business component. And work performed outside the United States is excluded regardless of who employs the staff, which matters for practices using offshore drafting and modelling support.

Case outcomes for design firms have gone both ways on facts that look similar from the outside, which is the honest reason to read the agreements rather than to rely on a rule of thumb. Our note on funded research and contract clauses works through what the courts have actually looked at.

The federal credit is rarely the whole number

Most states with a corporate income or franchise tax run their own R&D credit off a similar research base, each with a different rate, cap, carryforward, and refundability - and a federal number multiplied by a state rate is not a state credit. For a practice with offices in several states, where the qualified work was performed matters to both numbers.

§174A matters here too, and independently of the credit. It restored immediate expensing of domestic research and experimental costs for tax years beginning after December 31, 2024, on a definition broader than the four-part test - so design costs that fall outside the credit can still be §174A expenditures. A study that only chases the credit leaves that on the table.

State R&D credits Section 174A expensing

What we ask for first

The client agreements on your largest projects, in full
Project accounting by phase and task, not just by job
Option studies and superseded calculation sets
The states where the qualified work was actually performed
Which work was supported offshore, and how much

Summary only - the analysis is done per agreement and per business component, not per firm.

What a design-practice study can look like

A hypothetical scenario to show how the pieces fit together. It is not a quote, projection, or promise of results.

~60-person structural and civil practice

Illustrative

Technical payroll: $5.4M
Projects surviving the funded-research screen: ~40%
Share of those hours on genuinely uncertain work: ~25%
Estimated QRE: ~$540K
Illustrative federal credit: ≈ $32K-$54K

Plus the §174A deduction on domestic research and experimental costs, on a definition broader than the credit’s.

Illustrative only. Figures are hypothetical and rounded; the federal credit commonly works out to roughly 6-10% of QRE depending on method, filing history, and the §280C election. Note how much the two screens above remove: that is what an honest design-firm claim looks like. This is not a quote or a guarantee.

Don’t forget §174A

The deduction reaches further than the credit

IRC §174A restores immediate, full expensing of domestic research and experimental costs for tax years beginning after December 31, 2024, on a definition broader than the §41 four-part test. For a design practice, that gap is the point: costs that do not clear the credit can still be §174A expenditures.

Understand §174A

Engineering & architecture - frequently asked questions

Does architectural design qualify at all?

Parts of it can. The credit reaches work that is technological in nature - relying on the principles of engineering or the physical sciences - and that eliminates uncertainty through a process of experimentation. The engineering embedded in a building design can meet that. Aesthetic design is expressly outside it: §41(d)(3)(B) excludes purposes relating to style, taste, cosmetic, or seasonal design factors. The practical question on any project is which hours were resolving a technical unknown, and the honest answer is usually a minority of them.

Our contracts are fixed-fee. Does that settle the funding question?

No, and this is the most common mistake on this vertical. In 2024 the Eighth Circuit affirmed the Tax Court against a structural engineering firm on contracts that were fixed-price, holding that an obligation to comply with codes or to meet a general standard of care does not make payment contingent on the success of the research. A fixed fee is a good starting point; what has carried the day in other cases is express acceptance, rejection, and rework machinery. The analysis is done contract by contract - see how contract clauses decide who gets the credit.

We assign the drawings to the client. Have we given away substantial rights?

Not necessarily. Substantial rights do not mean sole rights: the Federal Circuit has held that a performer retains substantial rights where it may use the research results in its own business without paying for that use, even where those rights are not exclusive. In one Tax Court order involving an architectural design firm, client ownership of the delivered documents was not fatal because nothing prevented the firm from using the underlying research. That order is an unpublished summary-judgment ruling and is persuasive reasoning only, so the position has to rest on your own agreements rather than on the case.

Is energy modelling qualified research?

It can be, when the model is being used to resolve a genuine uncertainty rather than to document a decision already made. Calibrating a model, testing alternatives against it, and changing the design because the results said so is experimentation. Running a compliance model to demonstrate that a finished design meets a code baseline, or assembling a certification submission, is a documentation exercise and is generally outside the credit.

Does construction administration count?

Generally no. Once the design is settled and the component is in production - in construction, here - the work on it falls under the §41(d)(4)(A) exclusion for research after commercial production, and routine submittal review, RFIs, and site observation are not experimentation in any event. A genuine redesign forced by a condition discovered in the field can start a new business-component question, and it is that redesign, not the administration around it, that would be examined.

What about pass-through consultants the client reimbursed?

Watch the direction of the money. Contract research counts at 65% of amounts paid for qualified research performed on your behalf, and the agreement has to have been entered into before the research and to require you to bear the expense even if the research is not successful. Where a consultant is billed through to the client at cost, you generally have not borne the expense, and the deduction and the credit follow the party that did. More in qualified research expenses.

We use offshore drafting and modelling support. Does that count?

No. Only qualified research performed in the United States, Puerto Rico, or a U.S. possession is eligible; research conducted elsewhere is excluded under §41(d)(4)(F) regardless of who employs the staff or where the invoice is paid. It is one of the most common silent overstatements we see in professional-services claims.

Does §174A apply to a design practice?

It can, and its reach is wider than the credit’s. §174A restores immediate expensing of domestic research and experimental expenditures for tax years beginning after December 31, 2024, on a definition that is broader than the §41 four-part test. Costs that fall outside the credit can still be §174A expenditures, so the two populations should be identified separately rather than assumed to be the same - see our Section 174A guide.

Funded research: how your contracts decide who gets the credit

The risk and rights prongs worked through with the cases, including the design-firm outcomes on both sides.

Also relevant

Section 174A expensing

The deduction side, which reaches design costs the credit does not, for tax years beginning after 2024.

Find the hours that would actually survive a review

Tell us what your practice designs and how the work is contracted, and we’ll separate the genuinely uncertain hours from the documentation - screening every agreement for funded research, and capturing §174A - reviewed and finalized by R&D experts and backed by Audit Protection. Contact us for pricing tailored to your study.

[email protected] We typically reply within one business day.

Get your free credit estimate Talk to our team