There has been an important development for employers that offer a Dependent Care Assistance Program (DCAP). On August 11, 2026, the U.S. Department of the Treasury and the IRS released proposed regulations addressing the nondiscrimination requirements that apply to DCAPs under Internal Revenue Code §129. Among other things, the proposed regulations provide much needed clarification regarding how the 55% average benefits test is calculated.

And for employers that have struggled with this test, there is some potentially good news.

The Issue With the 55% Test

The 55% average benefits test is one of the nondiscrimination requirements that applies to a DCAP. Generally, the test requires the average benefits provided to non-highly compensated employees (NHCEs) to be at least 55% of the average benefits provided to highly compensated employees (HCEs).

For years, one of the questions surrounding this test has been who should be included in the calculation.

Under a more conservative interpretation, an employer could include all employees who are eligible for the DCAP in the average benefits calculation, even if an employee never elected or received a DCAP benefit. This can have a significant effect on the result because an employee who receives no benefit effectively contributes $0 to the average.

Consider an employer with 100 eligible NHCEs. If only 20 NHCEs participate in the DCAP, including all 100 eligible employees in the denominator could substantially reduce the NHCE average. This can make it much more difficult for an employer to satisfy the 55% test, particularly when participation among HCEs is significantly higher.

The proposed regulations provide clarification on this issue.

The New Proposed Calculation

Under proposed Treas. Reg. §1.129 2(d), the average benefit is generally calculated by taking the total dependent care assistance provided to the applicable group and dividing it by the number of employees in that group to whom dependent care assistance was actually provided during the plan year.

In other words, an employee who was eligible for the DCAP but received no dependent care assistance during the plan year generally is not included in the denominator for purposes of the average benefits calculation.

The proposed regulations specifically state that the 55% average benefits test is based exclusively on employees who receive more than zero benefits during the plan year.

Here's a simplified example.

An employer has 15 HCEs and 15 NHCEs who are eligible for the DCAP.

Eleven HCEs receive $7,500 each, for total HCE benefits of $82,500. Four NHCEs receive $7,500 each, for total NHCE benefits of $30,000.

Under the proposed approach, the employees who receive no benefits are not included in the average benefits calculation.

The HCE average is therefore $7,500, and the NHCE average is also $7,500.

The NHCE average is 100% of the HCE average, meaning the plan satisfies the 55% average benefits requirement.

The proposed regulations include examples demonstrating this approach.

Does This Mean Every DCAP Will Pass? No.

This clarification does not eliminate the DCAP nondiscrimination requirements or guarantee that a plan will pass testing.

The DCAP must still satisfy the applicable nondiscrimination requirements, including the eligibility test, contributions and benefits test, owner concentration test, and 55% average benefits test.

For example, an employer cannot simply exclude employees from the eligibility requirements to improve its testing results. The proposed regulations continue to require that the plan's eligibility classification be based on reasonable, objective business criteria and not discriminate in favor of HCEs.

The distinction is important: employees who are eligible but receive no DCAP benefit may be treated differently for purposes of the average benefits calculation than employees who are improperly excluded from eligibility.

Can Employers Use the New Method Now?

Yes. Although these are proposed regulations and are not yet final, Treasury and the IRS specifically state that taxpayers may rely on the proposed regulations for plan years beginning before the final regulations are published.

The proposed regulations are currently scheduled to apply to plan years beginning on or after the date the final regulations are published in the Federal Register. Until then, employers may rely on the proposed regulations if they choose to do so.

That distinction is important when discussing the new calculation with employers. The proposed regulations provide a new method that taxpayers may rely upon; they do not mean that every employer is required to immediately recalculate prior testing using the new methodology.

What About DCAPs That Already Failed?

This is where the new guidance may be particularly meaningful.

Employers that previously failed the 55% average benefits test may want to determine whether the results would change if the test were calculated using the methodology described in the proposed regulations.

However, a prior test should not simply be rerun without considering whether the underlying data remains appropriate for the testing year and whether the employer intends to rely on the proposed regulations.

If an employer wants to revisit a prior test, the appropriate testing data and applicable plan year facts should be reviewed to determine whether a retest is appropriate and what methodology should be applied.

For employers currently going through DCAP testing, this guidance may also provide an opportunity to consider the testing methodology before finalizing results.

Don't Forget the Bigger Picture

The 55% average benefits test is only one piece of DCAP nondiscrimination testing.

Employers should continue to consider the full set of applicable requirements, including whether the plan's eligibility provisions are nondiscriminatory, whether contributions and benefits discriminate in favor of HCEs, whether the owner concentration requirements are satisfied, and whether the plan is being operated according to its terms.

The new proposed regulations provide helpful clarification, but they do not eliminate the need for thoughtful DCAP nondiscrimination testing.

For employers that have historically struggled with the 55% test, however, the clarification regarding who is included in the average benefits calculation could make a significant difference.

If your DCAP previously failed the 55% average benefits test because nonparticipating eligible employees were included in the calculation, it may be worth taking another look at the results under the methodology described in the new proposed regulations.

As with any proposed regulatory guidance, employers should consider the facts and circumstances of their particular plan and consult with their legal or tax advisors regarding whether reliance on the proposed regulations is appropriate.

The DCAP 55% test just got a little clearer and that could be good news for some employers.