No More Counting Zeros: IRS Proposed Rules Make Dependent Care FSA Testing Easier and May Open the Door to the $7,500 Limit

Key Takeaways

• Proposed IRS regulations would significantly simplify dependent care assistance program (DCAP) nondiscrimination testing by excluding employees who receive no benefits from the 55% average benefits test. This participant-only approach should reduce testing failures and the need for employers to cap highly compensated employees’ elections.
• The new testing methodology will make it more practical for employers to adopt the increased $7,500 annual DCAP contribution limit and employers that previously maintained the $5,000 limit because of nondiscrimination concerns should consider modeling their results under the proposed rules before setting their 2027 plan design.
• The proposed regulations also provide employers with greater flexibility through a new numerical safe harbor for eligibility testing and a post-year-end correction mechanism for certain testing failures. Employers may rely on the proposed rules before they are finalized and should work with their Flexible Spending Account (FSA) vendors to revisit testing, highly compensated employee (HCE) caps, eligibility rules, and plan documents during the current open enrollment cycle.

For many companies, the end of the plan year brings an unwelcome headache: the DCAP fails its nondiscrimination testing, and HCEs learn that their elections must be capped or that part of their benefit will be taxable. That headache has also kept many employers from increasing the DCAP contribution limit from $5,000 to the higher $7,500 limit that took effect for 2026. Newly proposed regulations provide some relief for these issues.

On August 11, 2026, the Treasury Department and the Internal Revenue Service (IRS) published proposed regulations (REG-101355-26) that, while focused largely on employer contributions to the new Trump account contribution programs, provide long-awaited guidance on nondiscrimination testing for DCAPs under Internal Revenue Code Section 129. Except as noted, the nondiscrimination rules described in this article also apply to Trump account contribution programs, but we have focused our discussion on DCAPs because many companies already have them in place. For more on Trump account contribution programs, please see our related article here. Below, we explain what the DCAP testing change may mean for your company.

Why DCAP Testing Has Been a Problem

To keep the tax exclusion for HCEs, a DCAP must pass four tests: (1) a contributions and benefits test, (2) an eligibility test, (3) a test limiting benefits for more-than-5% owners to 25% of total benefits[1], and (4) the 55% average benefits test, which requires average benefits for non-highly compensated employees (NHCEs) to be at least 55% of the average for HCEs. If a plan fails, NHCEs keep their tax-free benefits, but HCEs do not.

The 55% test has caused most of the trouble for plan sponsors. Under the approach common in practice, eligible employees who did not enroll were counted as receiving $0. Because lower-paid employees often enroll at lower rates, those zeros dragged down the NHCE average, and many plans failed even when every employee had the same opportunity to participate. The usual fix was to cap HCE elections, often well below the legal maximum.

The Big Change: Only Participants Count

The proposed regulations would calculate average benefits for each group by dividing the total benefits provided to that group by the number of employees in that group who received benefits greater than zero. Non-participants would simply be disregarded.

For most employers, this should mean far fewer failures and far less need to cap HCE elections. The proposed regulations would also confirm that:

• Testing is done as of the last day of the plan year, counting anyone employed during the year who received benefits;
• Employees under age 21 or with less than one year of service, and certain collectively bargained employees, are excluded from the eligibility and average benefits tests; and
• For salary reduction benefits, employees earning less than $25,000 may still be disregarded.

Is It Time to Move to the $7,500 Limit?

For taxable years beginning after December 31, 2025, the annual DCAP limit increased from $5,000 to $7,500 ($3,750 for married individuals filing separately). Many companies have held off on adopting the higher limit, and for good reason. If a plan was already struggling to pass the 55% test at $5,000, allowing HCEs to elect up to $7,500 would only widen the gap between HCE and NHCE averages and make failure more likely. For those employers, raising the limit would have meant offering a benefit that HCEs could not actually use.

The proposed regulations change that calculation. Because only participants are counted, a plan where participating NHCEs and HCEs elect similar amounts may now pass comfortably, even at the higher limit. If your company has been thinking about raising its limit, now is a good time to work with your FSA vendor to take a fresh look and model 2027 testing.

A New Safe Harbor for Eligibility

Your eligibility rules must be reasonable and based on objective business criteria, such as job category, salaried or hourly status, or location. A classification that effectively lists employees by name will not qualify. The classification must also be nondiscriminatory under either a facts-and-circumstances test or a new numerical safe harbor. Under the safe harbor, the percentage of NHCEs eligible, divided by the percentage of HCEs eligible, must be at least 90%, reduced by 3/4 of a percentage point for each whole percentage point by which NHCEs exceed 60% of your workforce. For example, if 80% of your employees are NHCEs, the threshold is 75%. Most plans open to all employees on the same terms should pass easily.

A Safety Net: Fixing Failures After Year-End

If your plan fails the 55% test or the owner concentration test as of the last day of the plan year, the proposed regulations would let you treat the plan as passing by including the excess benefits in the affected HCEs’ (or owners’) taxable wages and reporting them by the Form W-2 deadline. In an IRS example, a plan that fails at 51.4% passes at 55.1% after the employer treats $500 of each HCE’s benefits as taxable wages. This gives you a defined correction path instead of relying only on mid-year election caps.

Can You Rely on the Proposed Rules Now?

Yes. Although the rules would formally apply only to plan years beginning on or after the date that final regulations are published, the IRS has indicated that employers may rely on them for plan years beginning before that date. Because most 2026 plan years are well underway, the real opportunity for most companies is the 2027 plan year, so the time to plan is during the current open enrollment cycle.

Your Action Checklist

1. Rerun your testing model using the participant-only method of testing.
2. Revisit HCE caps and consider whether they can be relaxed or removed.
3. Decide on the $7,500 limit based on modeled results.
4. Confirm eligibility rules meet the new safe harbor.
5. Update plan documents and communications to reflect any changes.
6. Watch for final regulations, which may differ from the proposal.

If you have questions about how these proposed rules affect your DCAP, or whether raising your limit makes sense, please contact the authors or your Foley & Lardner LLP relationship attorney.

[1] This prong of the DCAP testing does not apply to Trump account contribution programs.