A common question is whether failed experiments count for the federal R&D tax credit under Section 41. The short answer is that failed experiments can be evidence of a process of experimentation — the Treasury Regulations do not require the taxpayer to succeed — but failed work is not automatically qualified research. The underlying activity must still satisfy the four-part test, and documentation of failed alternatives can be valuable evidence. This page explains the framework in general terms. It is educational and is not individualized advice. For the foundational framework, see our page on qualified research.
Failure Can Be Evidence of Experimentation
The Treasury Regulations (§1.41-4) describe a process of experimentation as an evaluative process designed to evaluate one or more alternatives to achieve a result where the capability, method, or appropriate design is uncertain. Experimentation inherently involves tests that do not produce the desired result — that is how alternatives are evaluated and uncertainty is eliminated. A failed test of one alternative that informs the evaluation of another is exactly the kind of evaluative process the regulations describe. For more, see our page on process of experimentation.
The regulations also provide that a taxpayer need not succeed in developing or improving the business component. This means that a project that ultimately does not achieve its goal can still involve qualified research, provided the activity satisfied the four-part test while it was being conducted.
Failed Work Is Not Automatically Qualified
While failure can be consistent with experimentation, failed work is not automatically qualified research. The four-part test still applies: the activity must be undertaken for a permitted purpose, be technological in nature, be intended to eliminate uncertainty, and be conducted through a process of experimentation, all with respect to a business component. A failed project that did not involve a genuine technical uncertainty, that did not rely on hard-science or engineering principles, or that did not involve an evaluative process of alternatives may not constitute qualified research even though it failed.
The question is not whether the project failed, but whether the activity — while it was being conducted — satisfied the four-part test. Failure does not create qualification where the elements were not met, and success does not disqualify an activity that did meet them.
Iterative Testing and Failed Alternatives
Iterative testing — where a failed test of one alternative leads to the evaluation of another — is a common pattern in qualified research. The regulations describe the process as involving identifying uncertainty, identifying alternatives, and conducting an evaluative process. Failed alternatives are part of the evaluation: they help eliminate options and narrow the field. Documentation of which alternatives were tested, why they failed, and how the results informed the next steps can help demonstrate a genuine evaluative process. For more, see our page on trial and error.
Documentation Value of Failed Alternatives
Records of failed experiments can be particularly valuable evidence. They tend to show:
- that there was a genuine uncertainty (if the answer were known, there would be no need to test alternatives that fail);
- that the process was evaluative (the failure of one alternative informed the evaluation of others);
- that the work was directed at eliminating the uncertainty (the testing was designed to resolve the question);
- the sequence of the research (how one test informed the next).
Discarding or not recording failures can leave a gap in the story of how the uncertainty was addressed. For more on documentation practices, see our page on R&D tax credit documentation.
Hypothetical Example
Consider a manufacturing company that is developing a new coating formulation and is uncertain whether the coating can achieve a required corrosion-resistance standard. The company formulates and tests five alternative coating compositions. The first three fail to meet the standard, the fourth meets it marginally, and the fifth exceeds it. The failed tests of the first three alternatives are evidence of the evaluative process — they helped eliminate options and informed the direction of the research. The wages of the technicians, the materials consumed in all five tests (including the failed ones), and the test results may warrant review as qualified research expenses, provided the underlying activity constitutes qualified research.
This example is illustrative only and does not state that the costs definitely qualify. The key point is that the failures do not disqualify the costs; what matters is whether the activity satisfied the four-part test.
Key Takeaway
Failed experiments can be evidence of a process of experimentation — the Treasury Regulations do not require success — but failed work is not automatically qualified research. The four-part test still applies, and documentation of failed alternatives can be valuable evidence of an evaluative process. Because these determinations are fact-specific, professional review is appropriate before claiming the credit.