George v. Commissioner: What It Means for Your R&D Tax Credit

Agricultural innovation qualifies. But only if you can prove it happened.

Every year, food producers, agribusinesses, and livestock operations leave substantial R&D tax credits unclaimed not because their research doesn’t qualify, but because they assume that what happens in a barn or a broiler’s house doesn’t count as science. Under IRC §41, that assumption is wrong. And a landmark Tax Court decision issued in February 2026 makes that case better than anything in recent memory.

The Agribusiness R&D Assumption Most Companies Get Wrong

Section 41 of the Internal Revenue Code allows companies to claim a credit for qualified research expenses. The catch: research must meet a four-part test. It must involve a business component, rely on hard sciences, address genuine technical uncertainty, and follow a process of experimentation. At first glance, a poultry trial adjusting feed additives, testing new genetics, evaluating vaccine protocols can look like routine production work. Not research.

But Treasury Regulation §1.41-4 and decades of case law tell a more precise story. The question isn’t whether the work looks like lab science. The question is whether it resolves technical uncertainty through a systematic process of evaluation.

On a commercial farm, that standard is met more often than most companies realize.

The Condition Most Operations Overlook

The four-part test is where most agribusiness operators and many of their advisors stop reading. A company conducting feed trials or genetic line testing may be doing exactly what §41 contemplates: forming a hypothesis, running controlled comparisons, collecting production data, and drawing conclusions. But if no one documented that process as it happened, the credit doesn’t survive examination. That gap between qualifying activity and documented proof is where most agricultural R&D claims either succeed or collapse.

How George v. Commissioner Redefined R&D Credit Eligibility for Agriculture

GOMI, a major US poultry producer, claimed over $4.4 million in R&D tax credits for on-farm experiments aimed at improving bird health and growth. While the IRS initially denied the claims, arguing they were routine production rather than laboratory research, the Tax Court ultimately allowed four of GOMI’s seven trials.

The court recognized that real-world field variables like temperature and disease create uncertainty that sterile labs simply cannot replicate, making commercial-scale testing a valid form of research. Ultimately, the deciding factor between the allowed and disallowed claims came down to a single requirement: the quality of the documentation.

Three principles emerged from the decision that are already reshaping how IRS examiners approach agricultural R&D claims:

Agricultural innovation qualifies as research under §41 Farm-based experimentation to improve animal health, growth performance, or product uniformity meets the statutory definition when it addresses genuine technical uncertainty through a process of experimentation.
The pilot model doctrine extends to living production systems. Broilers raised in experimental flocks are pilot models under Section 174. Their associated costs, including feed, which comprised roughly 90% of GOMI’s claimed expenses, can qualify as research supply costs without requiring paired wage expenses.
Contemporaneous documentation controls the outcome Retroactive credit studies, testimony alone, and estimates without factual grounding are insufficient. The records created in the ordinary course of business flock reports, feed logs, mortality data, necropsy results are the evidence the Tax Court expects to see.

This matters well beyond the poultry sector. The decision follows JG Boswell Co. v. Commissioner (2022), which validated R&D credits for row crop farming. Together, these cases establish that agricultural innovation across animal production, crop science, and food processing is squarely within the scope of §41.

Two R&D Tax Credit Mistakes That Cost Agricultural Businesses Every Year

The George case illustrates two distinct failure modes, and both appear consistently in IRS examinations.

Mistake #1: Assuming the credit doesn’t apply

Producers performing systematic trials testing genetics, adjusting feed protocols, evaluating vaccines often never consider whether §41 applies to them. The research happened. The costs were incurred. The trials produced real data. But no credit was claimed, because no one asked the right question. GOMI itself did not claim credits on its originally filed returns for 2012 through 2014.

Mistake #2: Claiming the credit without building a contemporaneous record

GOMI claimed more than $4.4 million in credit. The Tax Court allowed a fraction of that not because the underlying science was invalid, but because three of the seven trials lacked the documentation to prove it. In one case, feed records contradicted the claim entirely, showing no dosage change had actually been made. In others, there were simply no flock-specific records linking the activity to the year at issue.

Both mistakes are avoidable. The George decision is, in effect, a blueprint for how to avoid them.

Why Documentation Can Make or Break Your Agricultural R&D Credit

The Tax Court’s opinion in George produced one ruling that deserves more attention than it has received: the near-complete rejection of the Cohan rule for R&D credit purposes.

The Cohan doctrine has historically allowed courts to estimate expenses when precise documentation is unavailable, provided there is a credible basis for the estimate. The Tax Court in George declined to extend that flexibility to base period QREs, stating plainly that it would not “wing it with an estimate ungrounded in the record.”

The rule is unforgiving but clear:

Records created in the normal course of business production logs, feed recipes, grower reports, mortality data are your strongest evidence
Testimony can support contemporaneous documentation, but it cannot substitute for it or contradict it
A credit study conducted after the fact, built entirely on interviews and estimates, will not hold up under examination

For agricultural businesses with significant R&D credit potential, the quality of your documentation at the time activities occur determines whether the credit survives. A well-supported claim built on real production records is far more defensible than a larger claim built on reconstruction.

One additional point from George: the court allowed GOMI’s supply expenses without requiring the company to also claim wage expenses. Businesses with complex payroll structures or research activities where tracking labor hours is impractical now have a clearer path to claiming supply-based QREs on their own.

How TaxDrone.AI and NTG Help Agricultural and Food Industry Companies Claim and Defend R&D Credits

The George decision established the legal framework. Applying it to a real company with dozens of trial flocks, years of production cycles, and thousands of line items across feed and supply costs is an entirely different challenge. That’s where TaxDrone.AI and NTG (National Tax Group) work together to bridge the gap between legal principles and defensible execution.

What George Requires	Without TaxDrone.AI + NTG	With TaxDrone.AI + NTG
Contemporaneous trial-level records	Manual reconstruction after year-end; relies on memory and estimates	Real-time capture linked to each flock, trial, and production cycle
Traceable supply cost connection	Cost categories allocated by percentage; no activity-level link	Each supply cost mapped to a specific qualified research activity
Four-part test applied to each trial	Applied broadly to the study as a whole	Applied trial-by-trial, with automatic exclusion of non-qualifying activities
Defensible sampling methodology	Informal; population never formally defined	Rev. Proc. 2011-42 compliant; population defined before sampling begins

Where TaxDrone.AI builds the record, NTG interprets it. Together, they address the two failure points that cost agricultural businesses the most: failing to recognize that research qualifies in the first place and failing to document it in a way that survives examination.

The Trials Have Already Run. Now It's Time to Document Them.

Most agribusinesses performing experimental work have already done science. The flocks were tested. The feed additives were compared. The genetic lines were evaluated against control. The only question is whether the production records support a credit and whether the documentation is strong enough to defend it.

If your company conducts systematic experimentation to improve animal performance, crop yield, product quality, or production efficiency, reviewing your eligibility under IRC §41 can significantly affect what you’re owed. Learn how our trial-level analysis approach works here.