Top-Heavy Plan

A plan is top-heavy for a plan year if, as of the determination date, the account balances of key employees exceed 60% of the account balances of all participants — the accrued benefit equivalent applies in a defined benefit plan. The rules live in IRC §416, headed "Special rules for top-heavy plans," with the definition itself at §416(g).

It is worth being precise about the grammar, because the phrase misleads. A "top-heavy plan" is not a species of plan alongside a profit-sharing plan or a money purchase pension plan. It is a temporary label a perfectly ordinary plan picks up in years when its assets are concentrated in the hands of the people who own or run the business, and drops again when they are not. The point of the rule is straightforward: a plan whose value has pooled at the top must give rank-and-file participants a floor contribution as the price of keeping its tax-favoured status.

It is measured on balances, not contributions. This is the single most common error, and it explains the pattern small employers actually experience. In a defined contribution plan the test is the ratio of key employees' account balances to all participants' balances; in a defined benefit plan it is the present value of accrued benefits. Because a long-tenured owner's accumulated balance grows year after year while newer employees start from zero, a plan that comfortably passed at launch can tip over a decade later without anything about the contribution design changing. Contribution percentages and pay levels are simply not what is being measured.

The determination date is the last day of the preceding plan year. So a calendar-year plan's 2026 status is fixed by balances as of 2025-12-31, and nothing an employer does during 2026 alters it. The obligation is known, and locked, before the year starts, which is genuinely useful for budgeting and genuinely unhelpful if you find out late.

Key employee is not highly compensated employee. These get conflated constantly and they are separate definitions in separate Code sections. Top-heavy testing uses only the key employee definition in IRC §416(i): an officer with compensation above an indexed threshold ($235,000 for 2026), or a more-than-5% owner with no compensation requirement whatsoever, or a more-than-1% owner with compensation above $150,000 — a figure that is written into the statute and is not indexed for inflation, so wage growth alone pulls steadily more owners into key status over time. A highly compensated employee, by contrast, is defined by IRC §414(q) with a different threshold and no officer prong at all.

The counting rules cut both ways. Distributions made in the one-year period ending on the determination date are added back into the calculation (five years for distributions made for a reason other than severance, death, or disability), so cashing an owner out does not immediately fix the ratio. In the other direction, the balances of former key employees are excluded entirely from both the numerator and the denominator, and employees with no service during the relevant period are excluded too. A plan can therefore move in or out of top-heavy status purely because a retired owner-officer drops out of the count. The test also runs on aggregation groups rather than plan by plan, so a standalone plan that looks fine can be top-heavy through the group it must be aggregated with.

The consequence. In a defined contribution plan, each non-key participant must receive an employer contribution of 3% of §415(c)(3) compensation — with one softening: if the highest contribution percentage received by any key employee is less than 3%, the non-key minimum drops to that lower percentage. A defined benefit plan's minimum is 2% of compensation per year of service, capped at 20%. Employer money already being contributed can generally count toward satisfying the minimum, so the practical cost is the gap rather than the whole 3%.

Two partial obsolescences, stated honestly. The top-heavy vesting requirement is largely redundant for defined contribution plans since the Pension Protection Act of 2006 tightened the general vesting maximums for employer contributions. And "super top-heavy", the 90% version of the test, is functionally dead: its only consequence was the combined limit under the old IRC §415(e), which the Economic Growth and Tax Relief Reconciliation Act of 2001 repealed for limitation years after 1999. The term survives in legacy plan documents, so readers do still encounter it. Escaping the status prospectively is usually done through a safe harbor 401(k) design, which carries an exemption under §416(g)(4)(H) — one that is fragile and evaluated year by year.

One SECURE 2.0 wrinkle is genuinely counterintuitive. Under §125, long-term part-time employees must be counted when determining whether the plan is top-heavy, but the employer may elect to exclude them from the top-heavy minimum contribution and vesting requirements. They influence the answer without necessarily receiving the consequence.

The plan's administrator totals account balances as of the determination date, splits them between key employees and everyone else, applies the add-back and exclusion rules, and compares the ratio with 60%. If it is over, the minimum contribution applies for the following plan year.

A hypothetical example. On 31 December 2025 a small firm's 401(k) holds $4,000,000 in total account balances, of which the owner and one officer, both key employees, hold $2,600,000. The ratio is 2,600,000 ÷ 4,000,000 = 65%, above the 60% threshold, so the plan is top-heavy for 2026. Each non-key participant must then receive an employer contribution of 3% of compensation, so a non-key employee earning $60,000 gets $1,800 for 2026, even if she deferred nothing herself.

The softening rule can change that materially. If the highest contribution percentage credited to any key employee for 2026 turns out to be only 2% of that person's compensation, the non-key minimum falls to 2% as well, and the same employee's floor is $1,200 instead. Note also what does not help: had the owner taken a $500,000 distribution in December 2025 hoping to shrink the numerator, that $500,000 would have been added back, and the ratio computed exactly as though the money were still in the plan. Figures are illustrative.

What the rule accomplishes

• Guarantees rank-and-file participants a real employer contribution in plans whose value has concentrated at the top.
• Locks the obligation in before the plan year begins, so the cost is knowable in advance rather than discovered at year-end.
• Where the employer is already contributing, existing contributions generally count toward the minimum, so the marginal cost is often small.

What makes it painful

• Status is driven by accumulated balances, so a successful long-running plan drifts into it with no change in design.
• The more-than-1% owner threshold is not indexed, so inflation quietly widens the key employee group every year.
• Non-key minimums are owed regardless of whether those employees contribute anything themselves, which small employers find hard to forecast.
• Adding discretionary profit sharing to a safe harbor plan can forfeit the exemption, so the two designs interact in ways that need costing.
• Legacy plan documents still refer to "super top-heavy," a test whose only real consequence was repealed decades ago.