Understanding ADP and ACP testing.

What the ADP and ACP tests are

The actual deferral percentage (ADP) test under IRC 401(k)(3) and the actual contribution percentage (ACP) test under IRC 401(m)(2) check that highly compensated employees (HCEs) do not benefit disproportionately from a 401(k) plan compared with non-highly compensated employees (NHCEs). The ADP test looks at elective deferrals; the ACP test looks at matching and employee after-tax contributions.

Safe harbor 401(k) plans that meet the contribution and notice requirements are deemed to satisfy the ADP test, and generally the ACP test for matching contributions, which is why many small plans use a safe harbor design.

Who counts as highly compensated

Under IRC 414(q), an employee is an HCE for a plan year if they were a more-than-5% owner in the current or preceding year, or had compensation above the threshold in the preceding year (the look-back year). The threshold is $160,000 for both 2025 and 2026 per IRS Notice 2025-67. A plan may elect to limit the compensation group to the top 20% of employees by pay. Family attribution rules apply to the ownership test, so spouses, children, parents and grandparents of a 5% owner are also treated as owners.

How the math works

1. Compute each eligible employee's ratio

   Actual deferral ratio (ADR) = elective deferrals for the year ÷ compensation. Eligible employees who did not defer count as 0%. Catch-up contributions are excluded.

2. Average by group

   Average the ratios separately for HCEs and NHCEs. Under the prior-year method the NHCE average comes from the previous plan year; under the current-year method, from the same year.

3. Find the HCE limit

   If the NHCE average is 2% or less, HCEs may be up to 2 times it. Between 2% and 8%, up to the NHCE average plus 2 points. Above 8%, up to 1.25 times it. Equivalently: the greater of 1.25 × NHCE, or the lesser of 2 × NHCE and NHCE + 2.

4. Compare

   The plan passes if the HCE average is at or below the limit. The ACP test uses the same limits with matching and after-tax contributions.

A worked example

Try your own figures in the ADP/ACP test estimator.

Correcting a failed test

When a test fails, the plan corrects under its terms and the regulations, most often by distributing excess contributions to HCEs, or by making qualified nonelective or matching contributions (QNECs or QMACs) for NHCEs. HCEs who are catch-up eligible may have excess amounts recharacterized as catch-ups up to their available limit.

Corrective distributions are calculated in two steps under Treas. Reg. 1.401(k)-2(b)(2): first, reduce the highest HCE ratios (leveling) to find the total excess; then distribute that total starting with the HCEs who deferred the most dollars. The home page corrective distribution walkthrough shows both steps.

Timing matters. Distributions made within 2½ months after the plan year ends (6 months for a plan with an eligible automatic contribution arrangement) avoid the 10% excise tax on the employer; for a calendar-year plan that is March 15. Failures must be corrected by the end of the following plan year to protect the plan's qualified status. See the IRS Fix-It Guide for the full rules.

Operational tips for TPAs

• Run a mid-year projection from year-to-date payroll so sponsors can adjust before year end.
• Validate the census first: wrong HCE flags and missing compensation are the most common cause of incorrect results.
• Put each plan's March 15 correction deadline on the calendar with an owner.
• Keep the test inputs, the result and the reviewer's approval together.

The nondiscrimination testing workflow shows how this runs in KinetiPlan.