What is nondiscrimination testing?
Nondiscrimination testing (NDT) is the annual process of verifying that a Section 125 cafeteria plan does not favor highly compensated individuals (HCIs) or key employees over the rest of the participating workforce. The testing compares how benefits are distributed across different employee groups and confirms that the plan's tax advantages are flowing equitably rather than concentrating at the top of the compensation structure.
A Section 125 plan that discriminates in favor of HCIs or allows key employees to receive more than their proportional share of tax-free benefits loses its qualified status for those individuals. When a plan loses qualified status for an HCI, that HCI's benefits become taxable income. The employer may owe back FICA and income tax withholding on those amounts.
The rules live in IRC Section 125(b) for the main cafeteria plan tests, IRC Section 125(c) for the key employee concentration test, and related provisions for FSA-specific testing. Because the IRS never finalized the proposed regulations it issued in 2007 (Prop. Treas. Reg. 1.125-7), employers operate under those proposed rules in good faith. The substantive requirements are well-established even though the formal regulatory guidance remains in proposed form.
Why the IRS requires it
The tax benefit of a Section 125 plan, specifically the ability to receive compensation in the form of pre-tax benefits rather than taxable cash wages, creates a potential for employers to skew the benefit toward highly compensated employees. Without testing requirements, an employer could structure a cafeteria plan that delivers maximum value to executives and minimal value to hourly workers while characterizing the arrangement as an employee benefit plan.
The nondiscrimination rules prevent this outcome. They require that the plan's benefits be distributed across the workforce in a way that reflects genuine availability to non-HCIs, not just technical access that functions as an HCI perk in practice.
For the vast majority of small and mid-size businesses with a genuine mix of employees, nondiscrimination testing is a straightforward annual confirmation rather than a difficult compliance hurdle. The testing problems tend to arise in two situations: plans with very small non-HCI workforces relative to the ownership or management group, and plans with long waiting periods or restrictive eligibility rules that effectively exclude most hourly workers.
The three Section 125 tests
A cafeteria plan must satisfy three distinct nondiscrimination tests under IRC Section 125:
- The Eligibility Test: The plan must not discriminate in favor of HCIs as to eligibility to participate.
- The Benefits and Contributions Test: The plan must not discriminate in favor of HCIs as to benefits or contributions available under the plan.
- The Key Employee Concentration Test: Nontaxable benefits provided to key employees cannot exceed 25% of all nontaxable benefits provided under the plan.
Each test applies independently. A plan can pass two tests and fail the third. FSAs have additional testing requirements on top of these three.
Test 1: The eligibility test
The eligibility test requires that the plan benefit a nondiscriminatory group of employees as to who is allowed to participate. If the plan's eligibility rules, whether through waiting periods, hours thresholds, or excluded employee classes, have the practical effect of excluding most non-HCI employees while covering HCIs, the plan fails.
How the test works:
The most commonly applied framework is the ratio percentage test. The calculation compares:
- The percentage of non-HCI employees eligible to participate
- The percentage of HCI employees eligible to participate
The ratio (non-HCI percentage divided by HCI percentage) must meet the applicable nondiscriminatory classification threshold. Under Section 410(b) principles applied to cafeteria plans through the proposed regulations, the ratio generally needs to be at least 70% or higher to satisfy the classification requirements, depending on the concentration of HCIs in the workforce.
Practical example:
| Employee group | Total employees | Eligible for Section 125 | Eligibility percentage |
|---|---|---|---|
| HCIs (officers, 5%+ shareholders, high earners) | 8 | 8 | 100% |
| Non-HCIs (all other employees) | 72 | 65 | 90.3% |
Ratio: 90.3% / 100% = 90.3%. This plan passes the eligibility test comfortably.
What causes eligibility test failures:
- Long waiting periods: A plan requiring 12 months of service before eligibility may exclude a disproportionate share of non-HCI hourly workers who turn over more frequently, while HCIs who have long tenure participate immediately.
- Restricted eligibility classes: Plans that cover only salaried employees or employees above a certain income threshold create eligibility discrimination if those classes are dominated by HCIs.
- Part-time exclusions: Excluding part-time employees from eligibility can create a discriminatory classification if part-time workers are predominantly non-HCI.
Safe design approach: Section 125 regulations limit waiting periods to 3 years of service. Most employers use 30 to 90 days. Shorter waiting periods reduce eligibility test risk. Broad eligibility classes that cover all W-2 employees on the same terms generally pass without issue.
Test 2: The benefits and contributions test
The benefits test verifies that the types and amounts of benefits available under the plan do not discriminate in favor of HCIs. If HCIs have access to a richer menu of benefits or higher employer contributions than non-HCIs, the plan may fail.
What the test measures:
The test compares whether HCIs receive a disproportionately higher level of employer-funded benefits or plan features relative to non-HCIs. In a typical Premium Only Plan where employees pay their own premiums through salary reduction, there is no employer contribution differential to test. The issue arises when the employer contributes different amounts to benefits for different employee groups, or when the benefit design effectively restricts the most valuable options to higher-paid employees.
Common benefits test issues:
- Tiered employer contributions: An employer contributing $500 per month toward health insurance for managers but $150 per month for hourly employees creates a contribution differential that must be analyzed.
- Opt-out cash arrangements: Some plans allow employees to receive cash compensation if they decline benefits. If HCIs disproportionately take the cash option while non-HCIs use the benefits, the plan's tax-advantaged structure may be under scrutiny.
- Executive-only benefits: Adding a benefit type available only to a class of employees dominated by HCIs creates benefits discrimination.
Practical guidance: For plans where both HCIs and non-HCIs elect and pay for benefits on the same terms through salary reduction, the benefits test is typically straightforward. The risk concentrates in plans with tiered employer contributions or flexible credit designs.
Test 3: The key employee concentration test
The key employee concentration test operates on a different metric from the eligibility and benefits tests. Under IRC Section 125(c), nontaxable benefits provided to key employees cannot exceed 25% of the total nontaxable benefits provided to all employees under the plan.
The test is a dollar concentration check, not a participation ratio check. If key employees are high earners with large benefit elections and the plan's non-HCI enrollment is low, the key employee dollar concentration can exceed 25% even when the eligibility and benefits tests pass.
How to calculate:
Sum the total pre-tax benefits for key employees and divide by the total pre-tax benefits for all participating employees. If the result exceeds 25%, the plan fails the concentration test for key employees.
Example:
| Employee group | Annual pre-tax benefits |
|---|---|
| 3 key employees | $54,000 ($18,000 each) |
| 47 other employees | $188,000 ($4,000 average) |
| Total | $242,000 |
Key employee concentration: $54,000 / $242,000 = 22.3%. This plan passes.
Now reduce non-key employee participation:
| Employee group | Annual pre-tax benefits |
|---|---|
| 3 key employees | $54,000 |
| 12 other employees who enrolled | $48,000 |
| Total | $102,000 |
Key employee concentration: $54,000 / $102,000 = 52.9%. This plan fails the concentration test.
The concentration test failure risk is highest in plans with low overall enrollment among non-key employees. Improving participation rates among the broader employee population is the most direct remedy.
Who counts as a highly compensated individual (HCI)?
The definition of HCI under IRC Section 125(e)(1) includes four categories:
1. Officers: Any officer of the employer, regardless of compensation level. Vice presidents, COOs, CFOs, and similar titled officers are HCIs even if their pay is below what you would typically consider high.
2. More-than-5% shareholders: Shareholders owning more than 5% of the voting power or value of all classes of employer stock. Attribution rules apply, so indirect ownership through family members may increase a person's attributed ownership above the 5% threshold.
3. Highly compensated employees: Employees who earned more than the applicable threshold in the prior year, generally tied to the Section 414(q) definition of highly compensated employee. For 2026, this is approximately $160,000 (confirm the current year threshold, as it is indexed annually). This also includes members of the top-paid group in some analyses.
4. Spouses and dependents of HCIs: Family members of the above categories are treated as HCIs themselves. A CFO's spouse who works as an HR coordinator is classified as an HCI.
One important note: The HCI definition for Section 125 is not identical to the highly compensated employee (HCE) definition used in 401(k) plan testing. Section 125 HCIs include officers regardless of pay and extend to family members. Applying the 401(k) HCE definition to Section 125 testing produces incorrect results.
Who counts as a key employee?
Key employees are defined under IRC Section 416(i) and used specifically for the concentration test (Test 3). The definition is different from the HCI definition used in the eligibility and benefits tests.
A key employee is anyone who, at any point during the plan year, is:
- An officer with annual compensation exceeding $220,000 (2026; indexed annually)
- A 5% owner of the employer (direct or attributed)
- A 1% owner with annual compensation exceeding $150,000
Note the distinction between HCIs and key employees: every officer is an HCI for Tests 1 and 2, but only officers earning over $220,000 are key employees for the concentration test. A company with many officers at moderate pay levels has more HCIs than key employees.
FSA and DCFSA testing
Health FSAs and Dependent Care FSAs carry additional nondiscrimination requirements beyond the three cafeteria plan tests.
Health FSA:
A health FSA must independently satisfy:
- The eligibility test (same as the cafeteria plan eligibility test)
- The benefits test (contributions cannot disproportionately favor HCIs)
- The key employee concentration test (same 25% cap)
Because the FSA is a component of the cafeteria plan, it must pass both the cafeteria plan-level tests and the FSA-specific tests.
Dependent Care FSA (DCFSA):
DCFSAs are governed by IRC Section 129, not Section 125, and have their own distinct nondiscrimination framework:
- Eligibility test: The DCFSA must benefit a nondiscriminatory group of employees.
- More-than-5% owners test: Benefits provided to employees who own more than 5% of the business cannot exceed 25% of total DCFSA benefits provided to all employees. This parallels the cafeteria plan concentration test but uses a different ownership threshold.
- 55% average benefits test: The average dependent care benefit provided to non-highly-compensated employees must be at least 55% of the average benefit provided to highly compensated employees. This is the DCFSA's most distinctive test and the one that most commonly creates compliance issues.
The 55% test in practice: If highly compensated employees average $4,500 per year in DCFSA benefits, the average benefit for non-HCEs must be at least $2,475 ($4,500 x 55%). If participation among non-HCEs is low, their average election will be dragged down by non-participants being counted at zero, and the plan may fail the test even if the non-HCEs who do participate are electing meaningful amounts.
DCFSA testing failures result in the excess benefits provided to HCEs becoming includable in income.
What happens when a plan fails
When a Section 125 plan fails a nondiscrimination test, the consequence is targeted. It falls on the people who benefited from the discrimination, not on the entire plan.
For the eligibility and benefits tests: The nontaxable benefits that would have been excludable from income are instead included in the gross income of the highly compensated individuals. Non-HCI employees who participated in the plan are unaffected. Their pre-tax elections remain valid and their benefits are not retroactively taxed.
For the key employee concentration test: If more than 25% of total nontaxable benefits went to key employees, the benefits exceeding the 25% threshold are included in the gross income of the key employees.
Tax consequences for affected individuals:
- The income that should have been excluded is added back to the HCI or key employee's taxable income for the year
- The employer may owe FICA on those amounts
- Income tax withholding obligations apply
- Potential penalties and interest if reporting and payment are delayed
A critical point: Because the failure consequences fall on specific individuals rather than invalidating the entire plan, the business continues operating normally. W-2 employees who were not HCIs or key employees see no change. The problem is contained, but it is still expensive for the affected individuals and creates additional payroll correction work for the employer.
Corrective action before year-end
Nondiscrimination testing failures discovered before the end of the plan year can be corrected without the full tax consequences of a failed plan. This is the primary reason to run tests mid-year rather than only at year-end.
Corrective action options:
Option 1: Increase non-HCI participation. If the concentration test is at risk because of low non-HCI enrollment, actively encouraging more non-HCI employees to elect benefits during a permitted enrollment window can shift the concentration ratio. This requires a qualifying reason to open enrollment mid-year under the plan's election change rules, or an administrative correction under a plan design change.
Option 2: Reduce HCI elections. HCIs can voluntarily reduce their benefit elections to bring the plan into compliance. This requires the HCI to experience a change in elections that is permissible under the plan's election change rules. Simply directing an HCI to reduce their election without a permissible trigger is not always straightforward.
Option 3: Amend the plan prospectively. Modifying eligibility rules or contribution structures mid-year to broaden non-HCI access is a prospective fix. It does not correct the period before the amendment but prevents continued deterioration toward year-end.
Option 4: Include excess amounts in HCI income prospectively. If correction is not practical, the employer can voluntarily include the excess benefits in HCI wages before year-end and withhold accordingly. This is a controlled failure resolution rather than an uncontrolled one.
Corrective action is significantly easier with guidance from an experienced benefits administrator. Summit Health Benefits provides nondiscrimination testing oversight as part of the plan administration relationship.
[[HUB_CONTACT_FORM]]
Plan design for compliance
Most nondiscrimination testing problems are preventable through thoughtful plan design. The structural choices made at setup determine how much testing risk the plan carries throughout its life.
Keep waiting periods short. The proposed regulations limit waiting periods to three years. In practice, most plans use 30 to 90 days. Shorter waiting periods mean more non-HCI employees become eligible faster, improving the ratio on the eligibility test without ongoing management.
Use broad eligibility classes. Plans that cover all W-2 employees on the same terms, with no carve-outs for employee category, salary level, or division, are structurally cleaner for eligibility testing than plans with multiple eligibility tiers.
Watch participation rates. Low non-HCI enrollment is the most common cause of concentration test risk. Employers who communicate benefits clearly, explain the paycheck math, and make enrollment easy tend to achieve higher participation rates and carry less concentration risk.
Be cautious with executive-only benefit features. Adding a benefit type, a higher contribution level, or a supplemental program available only to a management class dominated by HCIs creates both benefits test and concentration test exposure.
Run mid-year projections. Testing once at year-end gives you no room to correct. Running a projection at the mid-year point using year-to-date data allows corrective action while there is still time to change outcomes.
For S-Corp plans: Plans that correctly exclude more-than-2% S-Corp shareholders from participation often test more cleanly because the HCI ownership group is not inflating the plan's HCI benefit totals. See our companion article on Section 125 for S-Corp shareholders.
When to run testing
Section 125 nondiscrimination testing is based on plan year data and must be completed before the employer's tax return is filed for that plan year. For calendar year plans, that means the test must be completed and documented before the employer's federal tax return is filed, typically by April 15 or the extended due date.
The practical calendar most advisors recommend:
- Mid-year projection (June or July): Run testing on the first half's data to project where the plan will land at year-end. Identify any test at risk. Take corrective action while there is still time to affect the outcome.
- Year-end run (January, based on full year data): Finalize the test results using complete plan year data. Document the results and retain records with the plan file.
- Pre-filing confirmation: Confirm test results are on file before the employer's tax return is filed.
Testing should be documented in writing. "We think we pass" is not documentation. A compliant test includes the data source, the methodology, the HCI and key employee list, the calculation results for each test, and the conclusion.
Working with Summit Health Benefits
Summit Health Benefits is a fully managed benefits platform. Nondiscrimination testing is not an add-on or a separate engagement. It is built into the administration relationship from day one.
When we set up a Section 125 plan, we map the employer's full workforce, identify the HCI and key employee population, and structure eligibility language to minimize testing risk before the first plan year begins. During the plan year, we run mid-year projections and provide corrective guidance when any test is at risk of failing. At year-end, we finalize and document results before the employer's return is filed.
Our engagement includes the Section 125 plan document, adoption agreement, and SPD; annual nondiscrimination testing; IRS-limit amendments; payroll integration with major US payroll systems; year-end W-2 reconciliation; audit-defense file retention; and multi-language employee education and enrollment.
Administration is $35 per enrolled employee per month, billed from the employer's reduced FICA deposit. No setup fee. No long-term commitment. Employer net FICA recapture averages $91 to $136 per enrolled employee per month. After our fee, net employer savings are $56 to $101 per enrolled employee per month.
If you currently run a Section 125 plan and are not sure whether it has been tested, that is the place to start. Untested plans that have been running for multiple years may have accumulated failures requiring correction. The resolution is usually faster than the exposure.
This article is educational and not legal, tax, or ERISA advice. Section 125 nondiscrimination testing involves plan-specific and workforce-specific analysis. Engage a qualified benefits administrator or ERISA counsel for guidance on your specific plan.
[[FAQ_DATA]]
{
"items": [
{
"q": "What is the purpose of Section 125 nondiscrimination testing?",
"a": "Nondiscrimination testing ensures that a Section 125 cafeteria plan's tax advantages are distributed equitably across the workforce rather than concentrated among highly compensated individuals and key employees. The IRS requires annual testing to prevent employers from using pre-tax benefit plans primarily as a tax shelter for executives."
},
{
"q": "How often does nondiscrimination testing need to be run?",
"a": "Testing is required annually, based on plan year data, and must be completed before the employer's tax return is filed for that plan year. Most benefits advisors recommend running a mid-year projection (typically in June or July) and a final year-end test using complete annual data. Running only at year-end leaves no time for corrective action if a test is at risk."
},
{
"q": "What happens if our Section 125 plan fails nondiscrimination testing?",
"a": "When a plan fails a nondiscrimination test, the benefits that would have been excluded from income are instead included in the gross income of the highly compensated individuals or key employees who caused the failure. Non-HCI employees who participated in the plan are not affected. The employer may also face FICA obligations and income tax withholding requirements on the amounts that become taxable for HCIs."
},
{
"q": "What is the difference between an HCI and a key employee for Section 125 testing?",
"a": "Highly Compensated Individuals (HCIs) are used in the eligibility test and benefits test. The definition includes all officers, shareholders owning more than 5% of the company, employees above the compensation threshold (approximately $160,000 for 2026), and spouses and dependents of all three groups. Key employees, used in the concentration test, are defined more narrowly under IRC Section 416(i): officers earning over $220,000, 5% owners, and 1% owners with compensation exceeding $150,000. Every key employee is typically an HCI, but not every HCI is a key employee."
},
{
"q": "Can a small business fail nondiscrimination testing?",
"a": "Yes. The tests apply to all Section 125 plans regardless of employer size. Small businesses with a large ownership or management group relative to a small non-HCI workforce face more concentration test risk than larger employers with a broad employee population. However, most well-designed small business plans with broad eligibility and good employee communication pass testing without issue."
},
{
"q": "Do we need to run nondiscrimination testing if we only have a Premium Only Plan?",
"a": "Yes. The three Section 125 nondiscrimination tests apply to all cafeteria plans, including Premium Only Plans. A POP is a Section 125 plan, and the testing obligations follow automatically."
},
{
"q": "What records do we need to keep for nondiscrimination testing?",
"a": "Documentation should include the list of HCIs and key employees with the basis for each classification, the testing methodology used, the data inputs, the calculation results for each test, and the conclusion. Testing records should be retained in the plan file for at least six years, consistent with general ERISA and tax recordkeeping recommendations."
},
{
"q": "Is the DCFSA nondiscrimination test the same as the cafeteria plan test?",
"a": "No. Dependent Care FSAs are governed by IRC Section 129 and have their own distinct nondiscrimination rules, including the 55% average benefits test, which requires that the average DCFSA benefit for non-HCEs be at least 55% of the average benefit for HCEs. This test is separate from the three Section 125 cafeteria plan tests and applies independently."
}
]
}
[[/FAQ_DATA]]