The federal research tax credit under Internal Revenue Code Section 41 depends on the activities performed, the costs incurred, and the evidence connecting those costs to qualified research. Recent litigation emphasizes proof of technical uncertainty and experimentation at the business-component level. It does not establish a universal rule requiring one particular timekeeping system or excluding all retrospective evidence. Research deductions and the research credit also have different requirements, and legislation enacted in 2025 materially changed the treatment of domestic research expenditures.

The Foundations of Research Expenditure Law and the Limits of the Willits Analogy

Business commencement matters when distinguishing operating expenses from start-up expenditures, but the proposed lineage from Willits v. Commissioner to modern Section 41 substantiation requirements overstates the connection. Willits, T.C. Memo. 1979-294, predates both Section 195 and the federal research credit. It should not be described as a decision interpreting either provision or as the controlling source of a modern research-credit documentation mandate.

A more direct authority for pre-operational research deductions is Snow v. Commissioner, 416 U.S. 500 (1974). The Supreme Court interpreted the former Section 174 requirement that research be connected with a trade or business more broadly than Section 162’s requirement that the taxpayer already be carrying on a business. That deduction decision does not automatically establish eligibility for a Section 41 credit.

Section 41 generally requires research expenses incurred in carrying on the taxpayer’s trade or business. Section 41(b)(4) supplies a special rule for in-house research expenses of a taxpayer whose principal purpose is using the research results in the active conduct of a future trade or business. Pre-revenue status alone therefore does not determine credit eligibility. The nature of the expenditure, the taxpayer’s activities, and the applicable statutory exception must be examined separately.

The interaction of these provisions requires a year-specific analysis. A taxpayer should identify which entity performs the research, how it expects to use the results, and which expenses fall within each provision. A commercially viable product is not a universal prerequisite for research deductions or credits.

The Four-Part Test and the Process of Experimentation

Section 41 requires qualifying research expenditures, technological information, an intended application to a new or improved business component, and a qualifying process of experimentation. The permitted improvement concerns function, performance, reliability, or quality. The requirements apply separately to each business component, and the statutory exclusions remain applicable.

Technical Uncertainty

Technical uncertainty concerns the capability or method of developing or improving a business component, or its appropriate design, based on information available at the beginning of the research. Uncertainty about scheduling, customer preferences, or commercial demand is not enough. Existing engineering principles and software can be used in qualifying research; the taxpayer does not have to advance the knowledge of the entire industry.

Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, illustrates the difficulty of substantiating research credits for engineering design work. Project complexity, possible design revisions, and ordinary engineering activity do not by themselves prove that the statutory tests were met. The practical task is to connect an identified technical uncertainty to the alternatives evaluated and the work undertaken to evaluate them.

The Process of Experimentation

A qualifying process can include modeling, simulation, or systematic trial and error. A rigid laboratory protocol or a document explicitly labeled a hypothesis is not universally required. However, merely encountering a problem, choosing a design, or producing something new does not establish experimentation.

In Little Sandy Coal Co. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), the Seventh Circuit affirmed the denial of credits associated with vessel development. The novelty of the vessels did not substitute for evidence identifying experimental activities. The substantially-all requirement generally asks whether at least 80% of the relevant research activities constitute elements of experimentation, measured using cost or another consistently applied reasonable basis. Unsupported percentages cannot establish the threshold.

This business-component test must be distinguished from the separate rule that can treat all of an employee’s services as qualified when at least 80% are qualified services. An employee below that threshold can still have an eligible portion of wages. Neither rule makes every employee’s entire compensation automatically eligible.

Documentation Standards and the Limits of Estimation

The Cohan doctrine has not been categorically abolished for research-credit claims. Estimation requires an evidentiary foundation and cannot replace proof that research qualifies. Treasury Regulation Section 1.41-4(d) requires records sufficiently usable and detailed to substantiate the credit; it does not prescribe a single mandatory time-tracking format.

The Role of Time-Tracking

Project-level time records can strengthen a claim, especially when employees divide their work between experimental development, production, administration, and customer support. Design histories, test results, engineering correspondence, source-control records, and credible testimony may also help establish what happened and support a reasonable allocation. A retrospective study is only as persuasive as its supporting evidence.

The government’s 2024 filing in Kyocera AVX Components Corp. v. United States should not be treated as a judicial holding that time sheets are mandatory. A litigation motion states a party’s position; it does not, by itself, establish that the court adopted that position or dismissed the claim. The practical lesson is to support employee allocations with evidence of actual qualified services.

Evidentiary Requirements and Practical Documentation

The following comparison corrects the suggestion that courts replaced a permissive historical standard with a universal new recordkeeping mandate. The column labels describe an asserted contrast, not legally established eras.

Moore v. Commissioner, T.C. Memo. 2023-20, demonstrates the distinction between new product development and qualified services. The Tax Court accepted testimony that Nevco’s president and chief operating officer devoted substantial time to product development, but the evidence did not support an allocation to qualified research. It also rejected the asserted direct-supervision and direct-support grounds. His compensation was excluded from the credit calculation; the decision did not hold that all testimony or estimates are inadmissible.

The Funded Research Exclusion and the Impact of Local Law

Section 41 excludes research to the extent funded by another person or government. Treasury Regulation Section 1.41-4A(d) addresses both payment arrangements and substantial rights. Payments contingent on successful research results generally are not funding under this rule. If the performer retains no substantial rights, the research is fully funded. Where substantial rights remain, noncontingent funding may still reduce otherwise eligible expenses.

Contractual Analysis and Governing Law

In System Technologies, Inc. v. Commissioner, Docket No. 12211-21, an order dated January 3, 2025 denied the IRS’s motion for partial summary judgment. The court considered Indiana law, under which customers could recover payments if the company failed to deliver the agreed systems. That supported contingency of payment. The ruling addressed the funding issue presented by the motion; it was not a blanket determination that every claimed expense satisfied Section 41.

In Smith v. Commissioner, Docket No. 13382-17 and related cases, a March 7, 2025 order allowed consideration of foreign law while reserving its application to the contracts. That procedural ruling did not establish a general entitlement to credits whenever a contract was silent about intellectual property. The subsequent opinion, T.C. Memo. 2026-50, found substantial rights for four of six sample projects but not the other two. Payments were not contingent on research success; for the four projects with retained rights, potential credits were limited to qualifying research expenses exceeding funding.

These decisions support reviewing the complete contract and the governing jurisdiction’s law. Payment milestones, acceptance provisions, warranties, termination terms, ownership clauses, and retained-use rights must be considered together.

Rights and Risks for Government Contractors

Fixed-price and cost-reimbursement arrangements can allocate risk differently, but their labels alone do not decide credit eligibility. A termination-for-convenience clause does not automatically make a cost-plus-fixed-fee contract unfunded. The relevant question is whether payment depends on successful research and whether the contractor retains substantial rights. Applicable Federal Acquisition Regulation clauses, including FAR 52.227-14 where relevant, must be assessed in their contractual context. A possibility of losing profit or incurring unreimbursed costs is not a substitute for the required analysis.

Form 6765, Research Deductions, and Refund-Claim Review

Form 6765 and Business-Component Information

The December 2025 instructions make Section G optional for tax years beginning before 2026 and required for tax years beginning after 2025, subject to stated exceptions. Required filers generally list components in descending QRE order until reaching 80% of total QREs or 50 components, with remaining components aggregated. This filing threshold is distinct from the experimentation test.

Amended research-credit refund claims have separate information requirements. The IRS reviews whether a claim adequately identifies the business components, research activities, and expense categories. Administrative screening is not evidence of a new automated system deciding substantive credit eligibility, and a complete filing does not guarantee allowance.

The Section 174 and Section 174A Transition

For tax years beginning in 2022 through 2024, the general rule required five-year amortization of domestic specified research expenditures and fifteen-year amortization of foreign expenditures, using a midpoint convention. Public Law 119-21 changed the domestic rule: Section 174A generally permits immediate deductions for domestic research expenditures in tax years beginning after 2024, with an alternative capitalization election.

Foreign research generally remains subject to fifteen-year amortization under Section 174. Revenue Procedure 2025-28 addresses implementation and transition choices, including options for remaining domestic balances from 2022–2024 and special retroactive treatment for eligible small businesses. Election conditions and deadlines require individual review.

The scope of research deductions remains broader than Section 41 credit eligibility. Credit denial does not automatically determine the deduction treatment, and neither treatment follows merely from calling an expense research and development. The applicable tax year, domestic or foreign location, nature of the activity, and Section 280C coordination must be considered.

Strategic Implications and Future Outlook

Businesses should coordinate technical records, financial allocations, contract review, and tax-year-specific filing requirements. The goal is a defensible explanation of qualifying work and its costs, supported by records that can be traced back to the activities.

The Shrink-Back Rule

Where a business component does not satisfy the research requirements, the shrink-back rule applies the tests to its most significant subset of elements, continuing to smaller subsets until a qualifying subset is reached or no subset qualifies. A specialized cooling subsystem, for example, may require separate analysis even when the larger project includes extensive routine work. Narrowing the component does not eliminate the need to prove qualification and identify related expenses.