Cross testing is an IRS-approved retirement plan design strategy that can help business owners and highly compensated employees maximize profit-sharing contributions while satisfying nondiscrimination requirements. When employee demographics align, cross-tested profit-sharing plans offer greater flexibility, increased retirement savings potential, and tax-efficient plan design.
A profit sharing plan is an employer-funded feature of a defined contribution plan where company contributions are discretionary and tax-deductible. Many business owners pair profit sharing with a 401(k) to build retirement wealth faster. The problem? Under traditional contribution methods that apply a uniform percentage of pay to every participant, most owners never get close to the legal ceiling.
In 2026, the Section 415(c) annual addition limit is $72,000, subject to the 100% of compensation cap, and the 402(g) elective deferral limit is $24,500, with catch-up contributions of $8,000 for those age 50 and older, or $11,250 for participants ages 60–63.
Cross-testing changes that equation. It converts profit sharing contributions into projected retirement benefits at a normal retirement age — typically 65 — so that different employees can receive different contribution percentages and still pass IRS nondiscrimination testing.
The rest of this article explains when cross testing works best, how it works mechanically, and how PDG approaches plan design to maximize contributions without inflating plan costs.
Cross testing — also known as the comparability method or new comparability — is an IRS-approved approach to nondiscrimination testing under Internal Revenue Code §401(a)(4). It allows plan sponsors to prove that a profit sharing allocation does not improperly favor highly compensated employees by evaluating the plan on a benefits basis rather than a contribution basis.
In a typical DC plan, nondiscrimination testing compares contribution percentages between Highly Compensated Employees (HCEs) and non-highly compensated employees. Cross testing takes a fundamentally different approach: actuaries convert each participant’s annual company contribution into an equivalent benefit at retirement using interest rate assumptions, mortality tables, and a specified normal retirement age. The result is a projected benefit figure — sometimes called an Equivalent Benefit Accrual Rate (EBAR) — that normalizes for age differences.
This means different groups of eligible employees — owners, managers, staff — can receive very different contribution percentages as long as their projected retirement benefits are comparable. The method is most commonly applied to discretionary profit-sharing allocations layered on top of 401(k) deferrals and, where used, safe harbor contributions.
The core idea is the time value of money. Younger employees have 30–35 years of compounding ahead of them; older key employees and higher-paid owners may have only 5–10 years left to save. A small contribution rate for younger participants compounds into a substantial retirement benefit, while a larger contribution for an older owner has far less time to grow.
Under a flat 5% profit-sharing allocation, an owner earning $300,000 receives just $15,000 — nowhere near the $72,000 limit. With a cross-tested allocation, that same 60-year-old owner could receive a 20% contribution rate (approximately $60,000), while a 30-year-old staff member earning $50,000 receives 5% ($2,500). Because the staff member’s smaller allocation has decades to compound, the projected benefits at age 65 remain comparable, and the plan passes nondiscrimination testing.
Cross testing does not help every employer. Success depends on demographics, payroll mix, and the owner’s objectives.
Cross testing can work for smaller firms with 5–10 eligible employees, but it often delivers the most value for employers with 12–50 eligible employees where there is a distinct owner/key versus staff population. Larger groups give more statistical stability in rate group testing.
Cross testing is most effective when owners are significantly older than most employees — if that gap doesn’t exist, the advantage shrinks.
The calculations are complex, but the logic is straightforward. Here is the conceptual sequence:
This is the mechanism that allows the plan to be tested on a benefits basis even though it is a defined contribution plan — using projected retirement benefits for nondiscrimination testing, not current dollar amounts.
The age-weighted method is related but distinct: it uses a formula that inherently assigns larger contributions based on age and salary. New comparability instead groups employees into different classifications and relies on cross-testing to prove nondiscrimination, offering more precise control over who receives a larger contribution.
Cross-tested allocations must still satisfy the IRS general nondiscrimination regulations under §401(a)(4), including the gateway test and rate group testing. Gateway contributions must be made before cross-testing can occur.
The gateway test in concrete terms: each non-highly compensated employee must receive at least the lesser of:
For example, if the highest HCE allocation percentage is 18%, one-third equals 6% — but because 5% is the lower figure, NHCEs need at least 5%. Safe harbor nonelective contributions (such as 3%) and qualified nonelective contributions count toward this gateway requirement.
If the plan fails these tests, the employer typically must increase NHCE allocations, redesign employee groups, or pursue a plan amendment prospectively. This is why proactive annual review with a specialist like PDG matters — corrections are far cheaper than retroactive fixes.
Plan design is where the profit sharing leverage is created. Here are the common allocation approaches plan sponsors compare:
| Method | How It Works | Flexibility |
| Pro rata | Same percentage of pay for all eligible employees | Low |
| Integrated (permitted disparity) | Accounts for Social Security benefits by allowing higher allocations on pay above the taxable wage base | Moderate |
| Age-weighted method | Allocates based on age and salary — older participants get more | Moderate–High |
| New comparability (cross-tested) | Different groups receive different contribution percentages; tested via projected benefits | Highest |
Pro-rata methods distribute contributions equally among all employees, which limits owner benefit. The comparability method is often preferred when there is a targeted owner or partner group whose allocations need to be maximized.
Employers define allocation groups based on bona fide business criteria: ownership status, officer status, department, location, or full-time versus part-time. These classifications must be consistently applied and documented to satisfy IRS scrutiny. Employers can adjust contribution formulas annually under cross-tested plans based on profits, giving business owners meaningful control over company contributions each plan year.
The balancing act: maximize contributions for higher-paid owners, control total plan costs, meet gateway requirements, and keep allocations defensible. For 2026, an HCE earns over $160,000 — a threshold that determines which participants fall into the highly compensated group for testing purposes.
Cross-tested profit sharing rarely exists in isolation. It is usually layered on top of a 401(k) plan with employee salary deferrals and, often, safe harbor contributions. Safe harbor plans eliminate ADP testing headaches for deferrals — the ADP test compares deferral percentages between HCEs and NHCEs — but the cross-tested profit sharing piece still must pass 401(a)(4) nondiscrimination testing each year.
The strategic role of safe harbor: a 3% safe harbor nonelective contribution counts toward the gateway requirement. If NHCEs need 5% to pass the gateway test, the safe harbor covers 3%, and the employer only needs to add 2% in profit sharing to clear the floor.
Since 2003, The Pension Design Group has specialized in customized, full-service retirement plan administration for businesses that want to use their defined contribution and 401(k) plans strategically. PDG collaborates with the business owner, the company’s financial advisor, and the accountant to design a plan that attracts and retains top talent while meeting retirement goals and objectives.
Learn more about PDG’s approach or schedule a consultation.
Cross testing is most effective when business owners want to push toward maximum allowable contributions and employee demographics support the strategy. If you are considering this approach, here is what to gather before meeting with a specialist:
The right plan design can transform a standard 401(k) into a powerful retirement and tax-planning tool. A defined benefit plan or cash balance plan can be layered on top for even greater accumulation, though combined limits and coverage rules grow more complex. Decisions made for the 2026 plan year can have decades-long impact on owners’ retirement security.
Contact The Pension Design Group to model whether cross testing, an age-weighted method, or another type of plan such as Cash Balance is the most efficient path for your situation.
No. Cross testing affects employer profit sharing allocations and nondiscrimination testing, not individual elective deferral limits. In 2026, employees can still defer up to $24,500 under Section 402(g), plus catch-up contributions of $8,000 if age 50 or older, subject to the terms in the plan document. The contribution rate for deferrals is entirely separate from the cross-tested allocation.
Cross-tested allocations require specific plan document language. Changes are generally made prospectively for plan years via a plan amendment, and retroactive redesign is limited. Anti-cutback protections and IRS timing rules apply, so most employers adopt cross-testing effective for the next full plan year.
Cross-tested plans usually involve additional TPA and actuarial work each year for rate group testing and gateway verification. However, for many business owners, the incremental administrative cost is small compared to the thousands of additional deductible dollars directed to key employees and higher-paid owners through the cross-tested allocation.
Common corrections include increasing NHCE profit sharing for that year, reclassifying allocation groups, or making qualified nonelective contributions. In some cases, the plan can be deemed satisfied by adjusting the allocation formula before the filing deadline. Proactive mid-year modeling with a firm like The Pension Design Group dramatically reduces the risk of failure.
Yes. Many closely held businesses successfully combine a cross-tested defined contribution plan with a cash balance or other defined benefit plan to push total retirement savings even higher. However, combined contribution limits, deduction limits, and coverage rules become significantly more complex, making specialized plan design and ongoing administration essential.
Newsletter Signup