Section 129 DCAP Nondiscrimination Testing: New IRS Proposed Regulations Resolve Decades of Ambiguity

On August 11, 2026, the Treasury Department and Internal Revenue Service published proposed regulations that, for the first time, provide comprehensive regulatory guidance on how to perform nondiscrimination testing for dependent care assistance programs (DCAPs) under Section 129 of the Internal Revenue Code. While the proposed regulations were prompted by the need to implement the new Section 128 contribution rules for Trump accounts under the One, Big, Beautiful Bill Act, they also address longstanding ambiguities in Section 129 testing that employers and practitioners have struggled with for decades.

Link to The Section 129 Nondiscrimination Framework: A Refresher The Section 129 Nondiscrimination Framework: A Refresher

Section 129 allows employees to exclude from gross income up to $7,500 per year ($3,750 for married individuals filing separately) in employer-provided dependent care assistance, provided the DCAP satisfies four nondiscrimination requirements. If a DCAP fails any of these tests, the tax-free treatment is lost for highly compensated employees (HCEs), but non-highly compensated employees (NHCEs) continue to receive the exclusion.

The four tests are:

1. Contributions and Benefits Test. The plan’s contributions and benefits must not discriminate in favor of HCEs.
2. Eligibility Test. The plan must benefit employees who qualify under a classification established by the employer that is not discriminatory in favor of HCEs or their dependents.
3. Owner Concentration Test. No more than 25% of the amounts paid by the employer for dependent care assistance during the year may be provided to the class of individuals who are 5% or more shareholders or owners (or their spouses or dependents).
4. Average Benefits Test. The average benefits provided to NHCEs under all DCAPs of the employer must be at least 55% of the average benefits provided to HCEs.

Despite Section 129 being on the books since 1981, the IRS had never issued comprehensive regulations under this provision. The statute is short and high-level, and little IRS guidance existed. This left employers and practitioners operating in a regulatory vacuum for over four decades. The following are the issues that have most commonly vexed employers and their advisors.

• Low NHCE Participation Rates. Perhaps the most common practical problem with DCAP nondiscrimination testing has nothing to do with plan design. It is the reality that NHCEs participate in DCAPs at far lower rates than HCEs. DCAPs are funded through salary reduction, and lower-paid employees often cannot afford to set aside pre-tax dollars for dependent care or may not have qualifying dependent care expenses. The result is that a plan can be completely neutral on its face, offering the same election opportunity to everyone, and still face nondiscrimination exposure simply because HCEs disproportionately elect to participate. Previously, no clear authority confirmed that equal access was sufficient to satisfy the contributions and benefits test when utilization was skewed.
• The 55% Average Benefits Test: Who Counts? The statute provides that average benefits to NHCEs must be at least 55% of average benefits to HCEs, but it was unclear who should be counted in the denominator. Because HCEs would lose their tax exclusion under the DCAP if the plan fails the nondiscrimination test, many employers adopted a conservative testing approach, including all eligible employees in the denominator. This drove down the NHCE average (since many eligible NHCEs elected zero benefits) and made the test significantly harder to pass. Other employers counted only participants. This produced very different results but left their programs at risk for nondiscrimination failures.
• The Eligibility Test: No Safe Harbor, No Benchmarks. The statute requires a classification “found by the Secretary not to be discriminatory,” but the Secretary had never actually issued the framework for making that determination. No safe harbor existed, and no numerical benchmark told employers whether their eligibility structures were compliant. Employers that excluded part-time workers, imposed service requirements, or limited the benefit to certain business units had no way to measure whether those exclusions crossed a line. Practitioners borrowed concepts from the qualified plan nondiscrimination rules under Section 410(b), but no authority confirmed that this analogy was appropriate.
• The Owner Concentration Test in Closely Held Businesses. The 25% owner concentration test has posed a particular challenge for closely held businesses. In a company with a small employee population, a single owner’s DCAP elections could easily exceed the 25% threshold, even where the plan is offered broadly.
• No Clear Correction Mechanism. Under prior law, there was no established procedure for correcting a nondiscrimination failure after year-end. If an employer discovered at year-end that its plan failed testing, the consequence was simply that HCE benefits lost their exclusion and had to be included in income and reported on Form W-2. This all-or-nothing outcome forced employers to conduct testing throughout the year in an attempt to identify and address problems before they became irreversible. Even so, the lack of a defined correction procedure meant that an employer that discovered a failure had no sanctioned way to fix it.

Link to What the Proposed Regulations Would Change What the Proposed Regulations Would Change

The proposed regulations would provide significant new clarity in several areas, directly addressing some of the pain points described above.

Contributions and Benefits Test — A Qualitative Standard
The proposed regulations would confirm that the contributions and benefits test is a qualitative inquiry focused on the terms of the plan, not a mathematical comparison of amounts actually received. A plan would satisfy this test if it provides benefits on the same terms for all eligible employees, even if HCEs and NHCEs ultimately receive different dollar amounts because of differing elections or utilization rates. This would be welcome clarity. Under the proposed rules, an employer offering the same salary reduction opportunity to all eligible employees could be confident that different utilization alone will not trigger a failure.

Eligibility Test — A New Safe Harbor and Facts-and-Circumstances Framework
This is perhaps the most significant aspect of the proposed regulation. The proposed regulations would establish a numerical safe harbor and a facts-and-circumstances test, similar to the nondiscriminatory classification test for qualified retirement plans under § 1.410(b)-4:

• Numerical safe harbor: A classification would automatically be treated as nondiscriminatory if the plan’s “ratio percentage” — the percentage of eligible NHCEs divided by the percentage of eligible HCEs — meets or exceeds a safe harbor threshold. That threshold starts at 90% and decreases by three-fourths of a percentage point for each whole percentage point by which the NHCE concentration percentage exceeds 60%.
• Facts-and-circumstances test: If the safe harbor is not met, the classification can still pass based on facts and circumstances. Reasonable classifications may include specified job categories, hourly vs. salaried status, geographic location, or other similar bona fide business criteria. When evaluating a classification, the IRS may consider the underlying business reason for the classification, the percentage of employees eligible under the plan, whether eligibility is representative of the employer’s workforce across salary ranges, and the gap between the plan’s ratio percentage and the safe harbor. Classifying employees by name will not be considered reasonable.

Previously, employers had no regulatory benchmarks for evaluating their eligibility classifications. The proposed safe harbor would align Section 129 with established nondiscrimination principles from the qualified plan world and would give employers a concrete, numerical standard to design around.

Average Benefits Test: Clarifying the Denominator
The proposed regulations would resolve the longstanding ambiguity about who counts in the average benefits calculation. Under the proposed rules, the average benefit for each group (HCEs and NHCEs) would equal the total dollar amount of benefits provided during the plan year to employees in that group, divided by the number of employees in that group who actually received benefits greater than zero. Employees who were eligible but did not elect or receive any benefits are not included in the denominator.

This is a meaningful clarification. Under prior practice, many employers included all eligible employees in the denominator, driving down the NHCE average and making the test harder to pass.

The proposed regulations would also provide that the test is measured as of the last day of the plan year. The calculation takes into account anyone employed on any day of the plan year who is not an excluded employee and who received benefits during the year. Additionally, for benefits provided through salary reduction, employers may disregard employees earning less than $25,000 entirely.

Excluded Employees
The proposed regulations would formalize the categories of employees who are excluded from the eligibility and average benefits tests. These excluded employees are: (1) employees under age 21 who have not completed one year of service, and (2) collectively bargained employees not participating in the DCAP where dependent care benefits were the subject of good-faith bargaining.

New Remediation Pathway
For the first time, the proposed regulations would create a formal correction mechanism for DCAPs. If the average benefits test or the owner concentration test fails as of the last day of the plan year, the employer may remediate by including the excess benefit amounts (or excess ownership concentration) in the income of affected HCEs (or principal shareholders/owners) and reporting them as wages on Form W-2 by the W-2 filing deadline. The proposed regulations provide specific formulas for calculating the excess amounts and allow reasonable allocation methods among affected individuals.

This proposed remediation pathway is significant. Previously, a nondiscrimination failure was essentially an after-the-fact disaster with no clear correction procedure. Under these proposed rules, employers would have a defined window and methodology to fix the problem.

Link to Connection to Section 128 Trump Account Contribution Programs Connection to Section 128 Trump Account Contribution Programs

Section 128, the new provision governing employer contributions to Trump accounts, borrows its nondiscrimination framework from Section 129.Trump account contribution programs, discussed further here, must satisfy three of the four Section 129 tests: the contributions and benefits test, the eligibility test, and the average benefits test. The owner concentration test does not apply under Section 128, because self-employed individuals (who are the primary targets of that test) are excluded from participating in Section 128 programs altogether.

The proposed DCAP regulations are largely parallel to the proposed Trump account regulations, meaning the new clarity applies equally to both types of programs. The proposed Trump account regulations also include a special pilot match safe harbor for Section 128, which would allow employers to disregard contributions that match the government’s $1,000 Section 6434 pilot program payments when performing the contributions-and-benefits and average benefits tests, provided the match is offered on the same terms to all non-excluded employees.

Link to Thompson Hine Takeaways Thompson Hine Takeaways

• These are proposed regulations and are not yet final. The IRS has scheduled a public hearing for October 15, 2026, and comments are due 45 days after publication in the Federal Register. Taxpayers may rely on the proposed regulations for plan years beginning before the date final regulations are published. This should allow employers some comfort in relying on these proposed regulations for the 2026 plan year.
• Review your testing methodology. If your testing vendor or TPA has been including all eligible employees rather than only actual recipients in the average benefits test denominator, the proposed regulations could change your test results significantly. Plans that previously failed may now pass, and plans designed around the more conservative methodology may have unnecessary design restrictions that could be relaxed.
• Remediation would be available. Under the proposed regulations, employers that discover a nondiscrimination failure after year-end would no longer face an all-or-nothing outcome. They could correct the problem by including excess amounts in HCE income by the W-2 deadline.
• Closely held employers should take special note. The owner concentration test remains a trap for businesses with few employees and actively participating owners. The proposed remediation pathway would provide a defined procedure for addressing a failure, but prevention through plan design and mid-year monitoring remains the better approach.
• Section 128 and Section 129 testing are intertwined. Employers implementing Trump account contribution programs alongside DCAPs will need to perform nondiscrimination testing for both programs. While the tests are parallel, there are differences. Most notably, Section 128 does not have an owner concentration test and does have a pilot match safe harbor.