The Health FSA Nondiscrimination Tests

Question: What are the health FSA nondiscrimination rules, and why do employers almost always pass the test?

Short Answer: Health FSAs are subject to the §105(h) nondiscrimination rules. The components of these rules are many and layered, which creates a degree of complexity best left to the TPA to perform on the employer’s behalf. Nonetheless, the ultimate result is passing grade for almost all employers. Unlike the dependent care FSA’s nondiscrimination 55% average benefits test, the health FSA nondiscrimination tests offer sufficient alternatives and low thresholds that make it nearly unheard of to fail.

Starting Point: The Cafeteria Plan Nondiscrimination Rules
A typical Section 125 cafeteria plan that includes a pre-tax premium component, health FSA, and dependent care FSA is subject to three sets of nondiscrimination tests applicable to employers of all sizes:

1. Section 125 Cafeteria Plan Testing

2. Section 105(h) Health FSA Testing

3. Section 129 Dependent Care FSA Testing

For more details: Newfront Section 125 Cafeteria Plans Guide

The Section 105(h) Health FSA Nondiscrimination Tests
The §105(h) nondiscrimination rules are designed to prevent discrimination in favor of highly compensated individuals (HCIs) as to eligibility and benefits.

Highly Compensated Individuals (HCIs)
The §105(h) rules take a significantly different approach for determining who is a highly compensated individual (HCI) when compared to the better-known definitions under §125 for cafeteria plans (highly compensated participant) and §129 for dependent care FSA (highly compensated employee).

For purposes of the §105(h) health FSA tests, an HCI is:

• One of the top five highest-paid officers;

• A shareholder who owns more than 10% of the value of stock; or

• Among the highest-paid 25% of all employees.

The Health FSA Nondiscrimination Test Components
There are two main components of the §105(h) nondiscrimination rules:

1. The Eligibility Test; and

2. The Benefits Test

The health FSA must pass both of these tests to satisfy the §105(h) nondiscrimination rules. Each test has multiple subcomponents to address.

Flowchart of the Health FSA Nondiscrimination Tests

Health FSA NDT Component #1: The Eligibility Test
The §105(h) rules set out two different ways to pass the health FSA Eligibility Test:

1. The Percentage Test; or

2. The Classification Test

Passing either of these tests satisfies the health FSA eligibility test.

Eligibility Test Subcomponent #1: The Percentage Test (Rarely Passed)
The Percentage Test offers two alternative ways to pass:

• The 70% Test: Under this approach, the health FSA must benefit at least 70% of all employees; or

• The 80% Test*: Under this approach, the health FSA must benefit at least 80% of all employees who are eligible to benefit under the health FSA.

*The 80% Test is available only where at least 70% of all employees are eligible to benefit under the health FSA. Some therefore refer to it as the “70%/80% Test”.

Employers are rarely able to pass either of these tests for the health FSA because a relatively small percentage of the population elects to participate in the health FSA. Nonetheless, this is not a concern because employers can still pass the much easier Classification Test.

Note: Employers often receive a summary overview of the Section 125 cafeteria plan nondiscrimination test results that lists every category and subcategory of testing as “Pass” or “Fail”. The Percentage Test components of the test results will often show “Fail” given the nature of health FSA participation. Again, this is not a concern. Almost all employers can pass the alternative Classification Test without issue. In other words, it is perfectly fine to fail both components of the Percentage Test!

Eligibility Test Subcomponent #2: The Classification Test (Almost Always Passed)
The employer’s health FSA satisfies the Classification Test if the classification of employees eligible for the health FSA does not discriminate in favor of HCIs. In this case, the rules point to the 401(k) and other retirement plan nondiscrimination rules to determine what qualifies as a nondiscriminatory classification.

Under those retirement-side rules that are incorporated by reference in the health FSA context, there are two components of the Classification Test:

• The Reasonable Classification Test; and

• The Nondiscriminatory Classification Test.

The health FSA must pass both of these tests to satisfy the Classification Test.

Classification Test Subcomponent #1: The Reasonable Classification Test
This component requires that the classification of employees eligible for the health FSA must be “reasonable” and “established under objective business criteria.” The rules refer to eligibility based on specified job categories, nature of compensation (i.e., salaried or hourly), and geographic location as examples of reasonable classifications. The rules specifically point to listing specific HCIs by name (or anything similar) as not being reasonable.

A few additional notes about this Reasonable Classification Test:

• Most employers will easily pass the Reasonable Classification Test because they simply make the health FSA available to all employees eligible for the major medical plan.

• Even if the employer has selective health FSA eligibility classes, the groups of employees eligible for the health FSA are usually set by common, industry norm standards that are specifically noted as reasonable in the rules (e.g., classes set by region).

• There is no quantitative analysis needed to perform the Reasonable Classification Test, so employers do not need to wait for the FSA TPA NDT results to determine compliance—it is simply a function of confirming that the plan does not have a discriminatory eligibility structure by design on its face.

• In sum, it is very easy to pass the Reasonable Classification Test!

Classification Test Subcomponent #2: The Nondiscriminatory Classification Test
Under the retirement-side rules that are incorporated by reference in the health FSA context, there are two components of the Nondiscriminatory Classification Test:

• The Safe Harbor Test; or

• The Facts and Circumstances Test.

Passing either of these tests satisfies the Nondiscriminatory Classification Test.

Nondiscriminatory Classification Test Subcomponent #1: The Safe Harbor Test
The health FSA passes the Safe Harbor Test if the plan’s “ratio percentage” is equal to or greater than the plan’s “safe harbor percentage”. The “ratio percentage” is determined by dividing the percentage of non-HCIs benefitting from the health FSA by the percentage of HCIs who benefit.

If these NDT rules have not been complicated enough yet, the “safe harbor percentage” is where the rubber really hits the road. The safe harbor percentage starts at 50%, then is reduced by ¾ of a percentage point for each whole percentage point by which the “non-HCI employee concentration percentage” exceeds 60%. The “non-HCI-employee concentration percentage” is the percentage of all employees who are non-HCIs.

How can anyone track this level of detail? The regulations provide a useful table (Treas. Reg. §1.410(b)-4(c)(4)(iv)) showing the applicable safe harbor (and unsafe harbor) percentage for any non-HCI employee concentration percentage. But more importantly, remember this is why it is crucial to work with an FSA TPA that provides NDT services. Vendors that specialize in this area are much better equipped to perform this testing than employers.

Again, this test revolves around a safe harbor percentage to compare to the plan’s ratio percentage. If the ratio percentage is higher than the safe harbor percentage, the health FSA passes the Safe Harbor Test—and thereby passes Nondiscriminatory Classification Test. If the ratio percentage is lower than the safe harbor percentage (uncommon, but it does occur), the next step is to confirm the health FSA passes the Nondiscriminatory Classification Test via the alternative Facts and Circumstances Test.

Nondiscriminatory Classification Test Subcomponent #2: The Facts and Circumstances Test
In most cases, the employer’s health FSA passes the Safe Harbor Test, which makes the Facts and Circumstances Test unnecessary. However, if the plan does not pass the Safe Harbor Test, the Facts and Circumstances Test is fortunately an easy-to-pass alternative.

There are two components to passing the Facts and Circumstances Test:

1. The Unsafe Harbor Test; and

2. The Factual Determination Test

The health FSA must pass both of these tests to satisfy the Facts and Circumstances Test.

The Facts and Circumstances Test Subcomponent #1: The Unsafe Harbor Test
The health FSA passes the Unsafe Harbor Test if the “ratio percentage” is equal to or greater than the “unsafe harbor percentage”. As with the Safe Harbor Test above, the “ratio percentage” is determined by dividing the percentage of non-HCIs benefitting from the health FSA by the percentage of HCIs who benefit. The regulations also provide a useful table (Treas. Reg. §1.410(b)-4(c)(4)(iv)) showing the applicable unsafe harbor percentage for any non-HCI employee concentration percentage.

The unsafe harbor percentage ranges from 40% at the low-end of non-HCI employee concentration percentages to 20% at the high end. While it can get rather convoluted keeping track of the various percentages, tests, and subtests at this point, keep in mind that employers should be working with an FSA TPA that provides NDT services and can perform the test with minimal employer involvement.

There is also some important good news associated with the Unsafe Harbor Test: It is exceedingly easy to pass! Virtually all health FSAs have a ratio percentage that easily exceeds the (very low) unsafe harbor percentage threshold. Accordingly, although the NDT rules are somewhat of a labyrinth of testing components, and getting to the Unsafe Harbor Test portion is a winding path journey that can cause many to lose sight of the forest among the trees, the bottom line is it is almost impossible to fail this Unsafe Harbor Test.

The Facts and Circumstances Test Subcomponent #2: The Factual Determination Test
At this point, the tests within the tests have piled up so fast you need wings to stay above them. Fortunately, this the last subtest component of the Eligibility Test—and it’s another easy pass for most plans.

A health FSA passes the Factual Determination Test if, based on all the relevant facts and circumstances, the eligibility classifications are nondiscriminatory. The regulations list the following factors as relevant here:

• The underlying business reason for the classifications

• The percentage of employees who benefit from the plan

• Whether the percentage benefiting in each salary range is representative of the general workforce

• The degree by which the plan’s ratio percentage exceeds the applicable unsafe harbor percentage

Given that almost all employers have generic health FSA classification criteria (e.g., full-time status, geographic region, eligibility for the health plan) with no intended bias toward HCIs, this Factual Determination Test should almost never present an issue. This is mostly a “check the box” test unless the employer as a very unusual health FSA eligibility structure.

Note: If the health FSA does not pass the Safe Harbor Test but passes the Unsafe Harbor Test, the TPA will generally flag the result as needing additional review to confirm the plan satisfies the Factual Determination Test. This is not a concern because the Factual Determination Test is best viewed as a formality. Almost all employers clearly satisfy this test with nearly universal nondiscriminatory health FSA eligibility conditions.

Health FSA NDT Component #2: The Benefits Test
The §105(h) rules set out two components to pass the health FSA Benefits Test:

1. The Discriminatory Benefits Test; and

2. The Discriminatory Operation Test

The health FSA must pass both of these tests to satisfy the Benefits Test.

Benefits Test Subcomponent #1: The Discriminatory Benefits Test
The Discriminatory Benefits Test prohibits the health FSA from offering any benefit to HCIs that is not also available to non-HCIs. Examples of plan designs that would not satisfy the Discriminatory Benefits Test include:

• Providing a more generous employer health FSA contribution for HCIs

• Imposing a reduced health FSA salary reduction contribution limit for non-HCIs

• Restricting the list of health FSA reimbursable qualified medical expenses eligible for non-HCIs

There is no quantitative analysis needed to perform the Discriminatory Benefits Test, so employers do not need to wait for the FSA TPA NDT results to determine compliance—it is simply a function of confirming that the plan does not have a discriminatory benefits structure by design on its face. The good news is that it is unheard of for employers to impose any of these discriminatory conditions.

Most employers will easily pass the Discriminatory Benefits Test because they simply offer the health FSA on the same terms to all HCIs and non-HCIs who are eligible. In sum, it is very easy to pass the Discriminatory Benefits Test!

Benefits Test Subcomponent #2: The Discriminatory Operation Test
The Discriminatory Operation Test is best viewed as an umbrella provision preventing health FSAs from discriminating against non-HCIs in operation. Whether a plan is operationally discriminatory is a facts and circumstances test based on each plan’s specific arrangement.

As with many aspects of the §105(h) nondiscrimination rules, the structure and purpose of the Nondiscriminatory Operation Test does not apply naturally in the health FSA context. Section 105(h) applies to all self-insured health plans, not just health FSAs. In the context of a self-insured major medical plan, the Discriminatory Operation Test imposes an important safeguard against employer plan design decisions that are clearly intended as a subterfuge of the other components of the rules. With respect to health FSAs, it is hard to imagine employers instigating a plan operation in the real world that challenges these limits.

The main concern with this requirement appears to be that the employer does not selectively establish, amend, or terminate the plan in a manner designed to benefit HCIs. For example, an employer would act in a manner that is operationally discriminatory if it established a health FSA for an HCE to have a specific expense reimbursed, then terminated the health FSA upon the HCI no longer needing it or a non-HCI expressing interest in making an election. Employers uniformly do not contemplate that type of approach, and therefore this test is essentially never an issue for a health FSA.

The limited guidance available for the Discriminatory Operation Test confirms that the health FSA is not considered discriminatory in operation merely because HCIs participating in the plan utilize plan benefits to a greater extent than non-HCIs. Therefore, unlike the concerns with the dependent care FSA, health FSAs do not have to worry about whether HCIs are disproportionately electing and benefiting from the arrangement.

What Happens if the Health FSA Fails the Section 105(h) Nondiscrimination Tests?
This is very unlikely to occur! If the FSA TPA provides a result suggesting that the health FSA is failing the §105(h) nondiscrimination tests, the first step is to review the data inputs. In almost all situations, this is merely the result of inaccuracies in the flow of information to the vendor performing the NDT.

If the health FSA actually fails the test (extremely unlikely!) it would need to be corrected by the end of plan year (e.g., by reducing HCI benefits or increasing non-HCI benefits). Otherwise, HCIs would be taxed on all or a portion of the benefits they received under the health FSA, referred to as the “excess reimbursement.” Calculating the excess reimbursement is complicated, and the method differs based on whether the failure was in the Eligibility Test or the Benefits Test. But again, the health FSA should almost never reach this point because they almost always pass the §105(h) nondiscrimination tests without issue.

Reminder: The Dependent Care FSA NDT is Much More Difficult!
Dependent Care FSAs are subject to different NDT rules under Section 129. The 55% average benefits test (ABT) that applies to the dependent care FSA is by far the most difficult to pass. Dependent care FSAs that are nondiscriminatory in their plan design are nonetheless frequently unable to pass the ABT based on the plan’s utilization rates.

Highly compensated employees (generally $160k and higher in 2026) tend to disproportionately participate at higher levels and make higher dependent care FSA elections, which is exactly what the ABT is designed to prevent. That makes the ABT hard to pass without adjustments to reduce highly compensate employee elections.

Furthermore, we anticipate that the One Big Beautiful Bill’s 2026 limit increase to $7,500, as well as the bill’s enhancements to the alternative child and dependent care tax credit that make it more attractive for non-HCEs, will further exacerbate the dependent care FSA testing challenges for employers.

For more details:

• The OBBB Dependent Care FSA Increase Could Backfire

• The Dependent Care FSA 55% Average Benefits Test

Summary
The Section 105(h) health FSA nondiscrimination tests are quite complex and laborious in their detailed scope and multilayered subtests. However, the good news is a) most FSA TPAs will perform these tests on the employer’s behalf, and b) it is exceptionally rare for an employer to have any issue passing these tests. Each area in the NDT that is difficult to pass has a much easier alternative that virtually all health FSAs can clear without issue. So while NDT season is understandably a pain point for many employers when it comes to the dependent care FSA, employers should rest easy at least knowing that each year they will almost certainly pass the health FSA NDT without any suspense.

Regardless, employers should work with their FSA TPA each year to perform the full suite of cafeteria plan nondiscrimination tests and keep a record of passing or making any required corrections.

Relevant Cites:

Treas. Reg. §1.105-11:
(c) Prohibited discrimination.
(1) In general. A self-insured medical reimbursement plan does not satisfy the requirements of section 105(h) and this paragraph for a plan year unless the plan satisfies subparagraphs (2) and (3) of this paragraph.
…
(2) Eligibility to participate.
(i) Percentage test. A plan satisfies the requirements of this subparagraph if it benefits—
(A) Seventy percent or more of all employees, or
(B) Eighty percent or more of all the employees who are eligible to benefit under the plan if 70 percent or more of all employees are eligible to benefit under the plan.
(ii) Classification test. A plan satisfies the requirements of this subparagraph if it benefits such employees as qualify under a classification of employees set up by the employer which is found by the Internal Revenue Service not to be discriminatory in favor of highly compensated individuals. In general, this determination will be made based upon the facts and circumstances of each case, applying the same standards as are applied under section 410(b)(1)(B) (relating to qualified pension, profit-sharing and stock bonus plans), without regard to the special rules in section 401(a)(5) concerning eligibility to participate.
…
(3) Nondiscriminatory benefits.
(i) In general.
(A) Benefits. In general, benefits subject to reimbursement under a plan must not discriminate in favor of highly compensated individuals. Plan benefits will not satisfy the requirements of this paragraph (c)(3)(i)(A) unless all the benefits provided for participants who are highly compensated individuals are provided for all other participants.

Disclaimer: The intent of this analysis is to provide the recipient with general information regarding the status of, and/or potential concerns related to, the recipient’s current employee benefits issues. This analysis does not necessarily fully address the recipient’s specific issue, and it should not be construed as, nor is it intended to provide, legal advice. Furthermore, this message does not establish an attorney-client relationship. Questions regarding specific issues should be addressed to the person(s) who provide legal advice to the recipient regarding employee benefits issues (e.g., the recipient’s general counsel or an attorney hired by the recipient who specializes in employee benefits law).