Highly Compensated Employee (HCE) 2026: Definition, Threshold, and Why It Matters for Your Plan

"Highly Compensated Employee" sounds like a simple compensation threshold question, but in practice it's one of the most consequential designations in qualified plan administration. HCE status drives §401(a)(4) nondiscrimination testing, §410(b) coverage, §401(k) ADP/ACP testing, and the gateway calculations for cross-tested profit sharing plans. Get HCE classification wrong and every test downstream produces wrong results, which means corrective contributions, failed audits, or lost deductions.

This article explains the §414(q) definition for 2026, the two-part test (ownership and compensation), the often-overlooked top-paid group election, the ownership attribution rules under §318, and how HCE status interacts with the separate "Key Employee" designation that drives §416 top-heavy testing. For the full 2026 limit landscape, see the pillar guide on 2026 retirement plan contribution limits.

Key Takeaways for 2026

• The HCE compensation threshold is $160,000 for 2026, unchanged from 2025 (§414(q) indexes in $5,000 steps, and this year's COLA didn't clear the bar). An employee is an HCE if they earned more than that in the prior year (2025) or owned more than 5% of the employer at any time in the current or prior year.
• The ownership prong catches more people than you'd think. §318 attribution treats the owner's spouse, children, grandchildren, and parents as owners too, so an owner's 22-year-old earning $15,000 is an HCE. Siblings do not attribute.
• HCE is not the same as Key Employee. HCE (§414(q)) drives §401(a)(4), §410(b), and §401(k) ADP/ACP testing. Key Employee (§416(i)) drives top-heavy testing and has a separate $235,000 officer threshold for 2026. Most small-business owners are both, but the populations aren't identical.
• The top-paid group election caps the HCE group at the top 20% of employees by pay (plus all 5% owners), which can turn a failing §410(b) coverage test into a passing one when a large share of the workforce earns above the threshold. It's a documented plan-document choice, not a year-by-year toggle.
• First-year hires can only be HCEs by ownership. A new hire earning $500,000 with no equity stake is an NHCE in year 1; there's no prior-year compensation with that employer yet.
• Get the classification wrong and every downstream test is wrong (coverage, general nondiscrimination, ADP/ACP, cross-testing gateways), which means corrective contributions, failed audits, or lost deductions. HCE status is recalculated every year.

What Is a Highly Compensated Employee (HCE)?

Under IRC §414(q), an employee is a Highly Compensated Employee if they satisfy either of two tests:

1. Ownership test: Owned more than 5% of the employer (directly or by attribution under §318) at any time during the current or prior plan year.
2. Compensation test: Received compensation from the employer in the prior year that exceeded the §414(q)(1)(B) threshold.

The tests are independent. An employee who fails both is an NHCE (non-highly compensated employee) for all testing purposes. An employee who passes either test is an HCE, full stop.

Note that the ownership test looks at ownership at any time during the current or prior year, while the compensation test looks at prior-year compensation only. This asymmetry matters in the first year of employment (new hires have no prior-year comp with the employer, so they can only be HCEs by ownership in Year 1).

The 2026 HCE Compensation Threshold

For 2026, the §414(q) compensation threshold is $160,000. This is unchanged from 2025, because §414(q) is indexed in $5,000 increments under §415(d) and the cost-of-living adjustment this year didn't clear that bar.

Look-Back Year Concept

HCE status for 2026 is based on compensation earned in 2025. An employee who earns $250,000 in 2026 but earned $150,000 in 2025 is an NHCE for 2026 testing purposes (assuming they're not a 5% owner). They become an HCE in 2027 based on their 2026 earnings.

This look-back structure is designed to prevent gaming: plan sponsors can't simply raise an employee's comp mid-year to influence current-year testing results. Prior-year comp is locked in by the time the plan year begins.

First-Year Hires

New employees have no prior-year compensation with the current employer. Under §414(q)(1)(B), they can only be HCEs in their first year if they satisfy the ownership test. A new hire earning $500,000 in their first year with no ownership stake is an NHCE for Year 1 testing, even though their current-year compensation is well above the threshold. They become an HCE in Year 2 based on their Year 1 earnings.

How the Threshold Is Indexed

The §414(q) threshold is adjusted annually under §415(d) in $5,000 increments. The IRS publishes the cost-of-living adjusted limits each fall via an IRS Notice, typically in October or November, effective January 1 of the following year.

The Ownership Test and §318 Attribution

The ownership prong is where most HCE determination mistakes happen. Direct ownership is straightforward: count the shares. But §414(q) pulls in the attribution rules of §318, which treat an employee as owning stock held by certain family members or entities.

Direct Ownership (>5%)

If an employee directly owns more than 5% of the employer's stock (by value or voting power), they're an HCE. Full stop. The compensation test is irrelevant: a 6% owner earning $40,000 is still an HCE.

For non-corporate entities, the "more than 5%" test applies to capital or profits interest.

§318 Attribution Rules

Under §318, an individual is treated as owning stock held by:

• Spouse (unless legally separated)
• Children, grandchildren, and parents (but not siblings, a common surprise)
• Partnerships of which they're a partner (in proportion to their interest)
• Estates and trusts in which they have a beneficial interest
• Corporations in which they own 50% or more (proportionally)

Attribution runs "upward" from family members and entities. It does not run between siblings. It does run from parent to child and from child to parent, but not from sibling to sibling.

Common Attribution Pitfalls

The classic HCE-by-attribution trap: the business owner's 22-year-old child works summers at the business for $15,000 per year. Because the child owns stock by attribution from the parent (attribution from a parent to a child is an upward attribution that applies), the child is a 5%+ owner for §414(q) purposes. The child is an HCE despite earning nowhere near $160,000.

This cascades: the plan treats the child as an HCE for §401(a)(4) general testing, for §410(b) coverage testing, and for ADP/ACP testing. If the plan assumed the child was an NHCE (easy mistake), the tests run wrong and the results are invalid.

Another common mistake: counting a sibling's ownership. If two brothers each own 50% of a business, they're both direct 5%+ owners and both HCEs on that basis. But their sibling's ownership doesn't attribute to them under §318. A brother who doesn't own any stock directly, and whose only "ownership" would be by attribution from his brother, is not an HCE-by-attribution on sibling grounds alone.

The Top-Paid Group Election

Under §414(q)(1)(B)(ii), the plan can elect to limit the HCE group to the top 20% of employees by compensation, plus any 5% owners, regardless of comp.

What the Election Does

Without the election, every employee earning above $160,000 is an HCE. With the election, the plan only treats the top-paid 20% of employees (measured by current-year compensation) as HCEs on the compensation prong. The ownership prong still applies: all 5% owners are HCEs regardless of the election.

Why Employers Use It

For employers with a workforce where a large fraction of employees earn above the HCE threshold, the top-paid group election reduces the HCE count and narrows the population that has to be tested against. This can be the difference between passing and failing coverage and nondiscrimination testing.

Worked Example: Tech Startup with Above-Threshold Salaries

Consider a 50-employee technology company where 30 employees earn above $160,000 (the §414(q) threshold for 2026). Without the top-paid group election, those 30 employees are HCEs on the compensation prong, plus the two co-founders who each own 40% of the company, for 32 HCEs in all.

With the top-paid group election, only the top 10 employees by compensation (20% of 50) are HCEs on the compensation prong; the two co-founders are HCEs on the ownership prong regardless, and as high earners they sit within that top 10. Total HCE count drops from 32 to 10.

For §410(b) coverage testing (which compares the percentage of NHCEs benefiting to the percentage of HCEs benefiting), the top-paid group election reframes the math: 40 NHCEs and 10 HCEs instead of 18 NHCEs and 32 HCEs. Much easier to pass.

How to Elect

The top-paid group election is a plan document provision. It must be elected in writing, typically in the plan document itself, and it generally applies for the entire plan year in which it's made. The election can be changed, but only prospectively and with proper plan amendment, not retroactively to fix a failed test.

A plan sponsor can't "toggle" the election year by year based on testing outcomes. The election structure is meant to be a deliberate, documented design choice.

Why HCE Status Drives Nondiscrimination Testing

Every nondiscrimination test in the qualified plan framework compares HCEs to NHCEs. Getting the classification wrong means every test runs wrong.

§401(a)(4) General Nondiscrimination

Under §401(a)(4), a plan's contributions or benefits cannot discriminate in favor of HCEs. The general test compares each HCE's allocation rate (or accrual rate, in a DB plan) to the NHCEs in the HCE's "rate group," defined as NHCEs with equal or greater rates.

For a cross-tested profit sharing plan, §401(a)(4) is typically tested on a benefits basis using the Equivalent Benefit Accrual Rate (EBAR). The statutory testing interest rate is 7.5% to 8.5%, with UP-1984 unisex mortality. HCE misclassification inflates or deflates the wrong side of the rate group comparison.

§410(b) Coverage Testing

§410(b) requires that the plan cover a sufficient percentage of NHCEs relative to HCEs. The ratio percentage test requires the NHCE coverage percentage to be at least 70% of the HCE coverage percentage.

If an employee is misclassified as NHCE when they're actually an HCE, the NHCE denominator is too large and the HCE numerator is too small. The ratio percentage looks artificially favorable, masking a failed test.

§401(k) ADP/ACP Testing

The Actual Deferral Percentage (ADP) test and Actual Contribution Percentage (ACP) test compare the average deferral or matching contribution rate of HCEs to NHCEs. HCE misclassification changes both averages and can flip a pass to a fail or vice versa.

If the plan uses a Safe Harbor 401(k) design, ADP/ACP testing is bypassed: the safe harbor is the HCE escape hatch. But all the other §401(a)(4), §410(b), and §416 tests still apply.

§416 Top-Heavy Testing

§416 top-heavy testing uses a different classification: Key Employee. Key Employees include 5%+ owners, >1% owners earning above $150,000, and officers earning above $235,000 for 2026 (up from $230,000 in 2025). The top-heavy ratio compares the present value of Key Employee benefits to all-employee benefits; if it exceeds 60%, the plan is top-heavy and must provide minimum contributions or accruals to non-Key employees.

HCE and Key Employee status overlap heavily but aren't identical. See the comparison below.

HCE vs. Key Employee: Similar But Not the Same

The two designations get confused constantly. Here's a side-by-side comparison:

Most small-business owners are both HCEs and Key Employees: they're 5% owners, which triggers both designations automatically. But an officer earning $180,000 who owns no stock is an HCE (compensation > $160,000) and not a Key Employee (officer comp < $235,000). Conversely, a non-officer earning $170,000 who owns no stock is an HCE but not a Key Employee.

The practical takeaway: §414(q) HCE classification drives most nondiscrimination tests. §416 Key Employee classification drives top-heavy testing specifically. Both classifications have to be determined each year, using the correct thresholds and attribution rules.

How HCE Designation Affects Plan Design

Because HCE status drives so much testing, plan design often revolves around managing HCE-to-NHCE ratios to produce favorable test outcomes.

Why Combo Plans (DB + PS) Are HCE-Efficient

In a small-business combo plan (DB + profit sharing, often with a 401(k)), the HCE-to-NHCE contribution ratio can be extreme: the owner (HCE) might get $250,000+ of combined DB+DC contributions while each NHCE gets $5,000 to $10,000. Despite this, the plan passes §401(a)(4) when tested on a benefits basis using cross-testing and permitted disparity, because the testing math accounts for the NHCEs' longer time horizon to retirement.

The HCE population is typically small in owner-dominated practices: one or two owners and a handful of staff employees. The testing math works precisely because the HCE group is narrow.

Safe Harbor 401(k) as an HCE Testing Escape Hatch

A Safe Harbor 401(k) design bypasses ADP/ACP testing entirely in exchange for a fully-vested employer contribution (either 3% nonelective to all eligible employees, or a matching contribution meeting statutory thresholds). This is the most common solution for small-business 401(k) plans with high HCE deferral rates that would otherwise fail ADP testing.

Safe harbor doesn't exempt the plan from §401(a)(4) or §410(b) testing, just from the ADP/ACP pieces.

Cross-Tested Profit Sharing: Using HCE Status Strategically

Cross-tested profit sharing plans use different allocation rates for different employee groups (often putting HCEs in one group and NHCEs in another, or further tiered groups). §401(a)(4) testing is run on a benefits basis using age-adjusted EBARs. Because younger NHCEs have longer accumulation horizons, a small contribution for them produces a high EBAR that "carries" the HCE rate group.

This is how cross-tested plans can allocate 7 to 8 times more to HCEs than NHCEs and still pass. The math relies on correct HCE/NHCE classification as the starting point.

For the deduction-limit side of these structures and the DB interaction, see the Defined Benefit Plan Contribution Limits (2026) article.

Common Mistakes in HCE Determination

1. Using current-year compensation instead of prior-year for the comp test. The §414(q) compensation prong looks at the prior year. Mixing current-year comp into the test is a classic mistake.
2. Missing family attribution for owner's children employed by the business. Children of 5%+ owners are HCEs-by-attribution regardless of their own compensation or direct ownership.
3. Forgetting the top-paid group election when it would reduce HCE count. Tech companies, law firms, medical practices, and other high-comp professional employers often benefit from the election but never elect it.
4. Failing to recalculate HCE status each year. HCE determination is an annual exercise. An employee's classification can change year to year as compensation and ownership shift.
5. Confusing HCE with Key Employee. They're related but distinct. HCE drives §401(a)(4), §410(b), and ADP/ACP. Key Employee drives §416 top-heavy.
6. Applying §318 attribution incorrectly to siblings. Sibling attribution does not apply under §318. Two brothers each owning 50% are both 5% owners directly; a third sibling who owns nothing isn't an HCE-by-sibling-attribution.
7. Missing the first-year hire exception. New hires have no prior-year comp with the employer, so they can only be HCEs in Year 1 via the ownership prong.

The Bottom Line

HCE classification is the load-bearing wall under every qualified plan nondiscrimination test. The 2026 threshold is $160,000, unchanged from 2025, and the look-back year is 2025 for 2026 plan-year testing. The ownership prong pulls in §318 attribution, which catches most misclassification errors. The top-paid group election is an under-used tool for high-comp employers. And HCE is not the same as Key Employee, even though the two overlap heavily for small-business owners.

Getting HCE classification right is the foundation. Everything downstream (coverage testing, general nondiscrimination, ADP/ACP, cross-testing) rests on it. If the classification is wrong, the tests are wrong, and the corrective contributions or lost deductions can be real.

If you're designing a plan, evaluating test results, or advising someone through an HCE question, try the Cash Balance Proposal Tool for a same-day model of a small-business plan with correct HCE classification and nondiscrimination testing applied, or contact us to discuss a specific situation.

For the broader 2026 limit context, see the pillar article on 2026 retirement plan contribution limits. For the DB side of small-business plan design and how HCE status interacts with §415(b) and §401(a)(17), see the Defined Benefit Plan Contribution Limits (2026) article.