Clarification for DCAP Nondiscrimination Rules

On August 11, 2026, the Department of the Treasury and Internal Revenue Service issued proposed regulations addressing nondiscrimination rules for Dependent Care Assistance Programs (DCAPs), as well as related rules for employer contributions to Trump Accounts. For DCAP sponsors, the most significant development is proposed clarification of the §129 nondiscrimination tests, including a more workable approach to the 55% average benefits test. Although the regulations are only proposed, the preamble permits taxpayers to rely on the proposed regulations until final regulations are published.

For DCAP sponsors, the proposal is significant because there were not previously implementing regulations for how Code §129 requirements should operate. The proposal confirms the four existing tests and, in several respects, formalizes testing approaches already used in practice. It also provides useful new clarification regarding the eligibility test and the 55% average benefits test and the consequences of a testing failure.

DCAP Nondiscrimination Framework

§129 rules allow DCAPs to provide up to $7,500 each calendar year in tax-favored qualifying daycare assistance, but §129 rules also prohibit DCAPs from favoring highly compensated employees (HCEs) on a tax-favored basis. A difference in eligibility rules, waiting periods, employer contributions or benefit coverage can all potentially cause issues under §129 nondiscrimination rules if the more generous offering or coverage favors HCEs.

For this purpose, HCEs include >5% owners and employees whose compensation exceeds a certain threshold (e.g., for 2026, employees who received $160,000 or more in 2025). To determine compliance, §129 requires a DCAP to satisfy four nondiscrimination tests. A DCAP must pass all four tests; if any of the four tests are failed and timely corrections are not made, all DCAP benefits elected by HCEs must be included in their taxable income.

Four Nondiscrimination Tests

The four §129 nondiscrimination tests are outlined below. The proposed regulations clarify how several of them should be applied, particularly the eligibility and 55% average benefits tests.

Contributions & Benefits Test

Contributions and benefits may not discriminate in favor of HCEs or their dependents.

In practice, this means similarly situated non-highly compensated employees (NHCEs) and HCEs should generally have access to the same DCAP contribution opportunity and benefit structure. The proposed rules clarify that “A plan that provides benefits on the same terms for all eligible employees satisfies this requirement. For example, a term that allows all NHCEs to reduce salary in the same amount as HCEs does not discriminate as to contributions or benefits.”

Eligibility Test

Eligibility to participate may not discriminate in favor of HCEs or their dependents.

The proposed rules confirm that eligibility classifications should be based on reasonable, bona fide business criteria, such as job category, hourly versus salaried status, geographic location, or similar bona fide business criteria. Further, if a plan uses eligibility classifications rather than offering the DCAP to all employees, the proposed rules indicate the employer must be able to support the classification under either the applicable facts and circumstances analysis or the ratio percentage test described below.

Facts and Circumstances Analysis

This is a subjective test under which the IRS would consider whether a plan is nondiscriminatory based on all the relevant facts and circumstances, including the underlying business reason for the classification; the percentage of the employer's employees eligible under the plan; whether the number of employees eligible under the plan in each salary range is representative of the number of employees in each salary range of the employer's workforce; and the difference between the plan's ratio percentage and the employer's safe harbor percentage.

Ratio Percentage

The proposed rules apply a ratio percentage test that is similar to benefit eligibility testing used for other benefit nondiscrimination rules such as §105(h). The plan's ratio percentage is determined by dividing the eligibility percentage of NHCEs by the eligibility percentage of HCEs. The safe harbor percentage is 90 percent, reduced by ¾ of a percentage point for each whole percentage point by which the NHCE concentration percentage exceeds 60 percent. The NHCE concentration percentage is determined by dividing the total number of NHCEs by the number of total employees.

Owner Concentration Test

55% Average Benefits Test

Average benefits provided to NHCEs must be at least 55% of the average benefits provided to HCEs.

This has historically been one of the most difficult DCAP tests to pass. Many followed the interpretation that all eligible employees had to be included, even if they chose not to participate. The proposed regulations clarify that the calculation only needs to include employees who participated in the DCAP. As a result, NHCEs who are eligible but do not elect DCAP benefits will not dilute the NHCE average benefit amount.

*For nondiscrimination testing, an employee or owner who was eligible or participated at any time during the year but terminated employment prior to year-end should generally be included in the testing. For purposes of the eligibility test and the average benefits test, the testing may exclude employees who are under age 21, employees who have <1 year of service, and employees who were not included in the DCAP and were covered by a collective bargaining agreement under which dependent care benefits were the subject of good faith bargaining.

Testing Failures

The proposed rules clarify that a testing failure does not disqualify all DCAP benefits. Instead, the correction generally focuses on the discriminatory portion of benefits provided to HCEs. HCEs must include the applicable taxable amount in income, and the employer should reflect that amount through appropriate Form W-2 reporting by the otherwise applicable deadline.

To avoid having to include all DCAP benefits in an HCE’s taxable income for a plan that otherwise fails any of the §129 discrimination tests, the employer can include only the discriminatory/excess benefit so long as it is done on the HCEs’ Form W-2 no later than January 31st following the end of the applicable plan year. This clarification is significant because prior guidance did not clearly provide this correction mechanism, and many practitioners interpreted the rules as requiring correction before the end of the plan year.

Summary

Employers sponsoring DCAPs should consider whether the proposed rules create an opportunity to revisit plan design, participation strategy, and testing assumptions for the current or next plan year. In particular, employers that previously limited DCAP availability because of 55% average benefits testing concerns may have more flexibility if they can rely on the participant-only averaging approach.

Employers should also coordinate with their third-party administrators and payroll teams to confirm how testing will be performed, how excludable employees will be identified, and how any HCE income inclusion will be reported if a test fails. Because the preamble permits reliance on the proposed regulations before they are finalized, employers may begin applying the clarified testing rules now.

If you have questions, please contact your North Risk Partners Risk Advisor. Don't have an advisor? No problem. We'll help you find one.

Find an Advisor

While every effort has been taken in compiling this information to ensure that its contents are totally accurate, neither the publisher nor the author can accept liability for any inaccuracies or changed circumstances of any information herein or for the consequences of any reliance placed upon it. This publication is distributed on the understanding that the publisher is not engaged in rendering legal, accounting, or other professional advice or services. Readers should always seek professional advice before entering into any commitments.

Posted in Compliance