George v. Commissioner: What Every R&D Taxpayer Needs to Know About This 2026 Tax Court Case

When people think about R&D tax credits, they tend to picture pharmaceutical labs, aerospace engineers, and software developers. Poultry farming does not usually make the list. That is part of what makes George v. Commissioner, T.C. Memo. 2026-10, so instructive.

In this case, the U.S. Tax Court examined whether large-scale poultry production activities qualified for the research credit under IRC Section 41. The court allowed some credits, disallowed others, and abated accuracy-related penalties. The outcome was a split decision, and the reasoning behind it carries lessons that apply to virtually every industry claiming the credit.

Background: What the Taxpayer Was Doing

The taxpayer was engaged in commercial poultry production at scale. They argued that certain production activities qualified as research under the 4-Part Test because they involved genuine technical uncertainty, a process of experimentation, and a technological purpose tied to improving yield, disease resistance, feed efficiency, or other measurable outcomes.

The IRS disagreed on a number of specific activities and argued that much of what the taxpayer described as experimental was simply ordinary production. The court had to work through the record activity by activity to determine what qualified and what did not.

What the Court Allowed

The court did allow credits for certain activities where the taxpayer was able to demonstrate that technical uncertainty existed at the outset of the work and that the production activities in question were structured as genuine experiments rather than standard commercial runs. Critically, these were activities where the taxpayer had documentation that identified what they were trying to learn, how they structured the test, and what they observed.

The court also found that some commercial production costs qualified because those production runs functioned as pilot models. When production is genuinely designed to resolve uncertainty rather than simply to produce product for sale, it can meet the standard. The key is that the research purpose has to be the driver, not a byproduct.

What the Court Disallowed

The disallowed activities had a common thread: the taxpayer could not show that technical uncertainty existed at the time the work began. In several instances, the court found that earlier activities had already resolved the relevant uncertainty, and the subsequent expenses were just ordinary production costs incurred after the answers were already known.

This is the later-failure trap that trips up a lot of taxpayers. If a batch of poultry does not perform as expected, or a crop fails, or a product does not work in the field, that failure after the fact does not mean uncertainty existed from the start. Uncertainty under Section 41 is evaluated at the beginning of the activity, not in hindsight.

Takeaway: A project that fails is not automatically a research project. Uncertainty must exist at the outset of the work, not be invented retroactively because something went wrong.

The Cohan Rule Problem

The court also addressed the taxpayer’s use of estimates to quantify qualified research expenses where precise records were not available. Courts have long recognized the Cohan rule, which allows a taxpayer to estimate expenses with reasonable support when exact records do not exist. But courts are increasingly skeptical of applying Cohan in the R&D credit context, and George v. Commissioner reflects that.

Where the taxpayer relied on estimates without solid evidentiary support for the underlying calculations, the court was not willing to credit those amounts. This is a significant warning for any taxpayer who is quantifying QREs through estimates or allocations without a clear methodology tied to actual records.

The Penalty Defense That Worked

One of the more encouraging aspects of the decision was the court’s treatment of accuracy-related penalties. The IRS had asserted penalties, but the court abated them on the basis that the taxpayer had relied in good faith on qualified advisers. The taxpayer had engaged competent professionals to analyze the credit, followed their guidance, and had a reasonable basis for the positions taken even where the court ultimately disagreed with them.

This is a meaningful reminder that penalty exposure and tax liability are two separate questions. Even when a credit claim is partially disallowed, a well-documented reliance on qualified professional advice can protect against the additional 20% accuracy-related penalty on top of the tax owed.

Takeaway: Penalties and tax are separate questions. Good-faith reliance on a qualified adviser is a real defense, and it held up in this case. Document your engagement with advisers and keep records of the analysis they provided.

Why This Case Matters Beyond Poultry

George v. Commissioner is not really about chickens. It is about the documentation infrastructure that every R&D taxpayer needs to have in place regardless of industry.

The activities that survived were the ones where the taxpayer could point to contemporaneous records showing what the uncertainty was, how the experiment was structured, and what happened. The activities that did not survive were the ones where the taxpayer was essentially asking the court to infer research from outcomes rather than demonstrate it from records.

That dynamic plays out in every industry. A pharmaceutical company that cannot show what technical uncertainty it was trying to resolve at the start of a trial. A manufacturer that cannot connect production runs to a defined experimental protocol. A software team that never documented the technical problems it was solving. They all face the same risk the George taxpayer faced.

What This Means for Your Documentation Practices

The lessons from this case are practical and applicable right now, regardless of where you are in your R&D credit cycle.

• Document uncertainty at the start, not at the end. Before an activity begins, capture what you do not know yet and what you are trying to find out. This does not need to be elaborate. A short description written at the time is worth far more than a detailed reconstruction written later.
• Define your experimental unit. What is the specific thing being tested? A batch, a formulation, a process parameter, a system configuration? The more precisely you can identify the unit of experimentation, the easier it is to defend.
• Track deviations from standard. When production or development goes differently than expected, note it. Those deviations are often the clearest evidence that genuine experimentation was occurring.
• Separate research costs from production costs carefully. If your claim includes costs incurred after you believe uncertainty was resolved, be prepared to explain why. The IRS and courts will scrutinize that boundary closely.
• Do not rely on Cohan if you can avoid it. Build your QRE calculation on actual records and a defensible methodology. Estimates are a last resort, not a first option.
• Document your adviser engagement. Keep records of the analysis your advisers performed and the basis for the positions they recommended. That documentation is your penalty defense if you ever need it.

Takeaway: The George case is a useful reminder that even industries that have to work harder to establish R&D qualification can succeed. But the foundation is always the same: contemporaneous records that show what the uncertainty was, what you did to resolve it, and how you know.

If you want to talk through how your documentation practices would hold up in an audit, reach out. We’d love to talk.

Jason Massie, CPA, Esq.

As the Founder and President of MASSIE R&D Tax Credits, Jason is responsible for the quality, consistency, and integrity of our services. His strategic vision and leadership drive innovation within the company, encouraging collaboration and ne...