Average Retirement Savings by Age — and the Rise of the 401(k) Millionaire
The Million Dollar Question: As of mid-2026, how many individual 401(k) accounts in the U.S. had reached $1 million or more, according to Fidelity Investments?
A) 269,000 B) 495,000 C) 769,000 D) 1.2 million
Read on for the answer.
Fidelity just reported a record number of seven-figure 401(k) accounts, growing faster in a single quarter than at any point since 2023. Days later, the Federal Reserve released its newest household survey, and it tells a very different story about the typical saver. Both are accurate. The trick is understanding how they fit together — and by the end of this one, you’ll know exactly where you actually sit.
What “retirement savings” actually means
“Retirement savings” usually means money sitting in a defined-contribution account — a 401(k), 403(b), 457, Thrift Savings Plan, or IRA — where the balance is whatever you and your employer put in, plus however the markets have treated it since. That’s different from a pension (a defined-benefit plan, where the employer promises a fixed payout regardless of markets) and different again from net worth generally, which also counts home equity, taxable brokerage accounts, and business ownership.
Two separate measurement systems report on this money, and they rarely agree with each other, which is where most “average retirement savings by age” headlines go wrong. The first is provider administrative data: Fidelity and Vanguard sit on tens of millions of real accounts and can report, in close to real time, what the mean balance is across everyone who already has an account with them. The second is the Federal Reserve’s Survey of Consumer Finances, a nationally representative household survey conducted roughly every three years, which asks a random sample of American families what they actually have — across every provider, including people with no account at all — and reports both the median and the mean. Fidelity’s numbers describe people who already save. The Fed’s numbers describe everyone.
Who actually has a seven-figure balance — and who doesn’t
Start with the population size, because it’s smaller than the headlines suggest. Bloomberg’s reporting on Fidelity’s data put the number of $1-million-plus 401(k) accounts at a record 769,000 as of the end of the second quarter of 2026 — up 19% in a single quarter, the largest quarterly jump since late 2023, driven in large part by the S&P 500’s roughly 15% gain over the same three months. That’s a real number, and it’s growing fast. It’s also a rounding error against the universe of U.S. retirement savers: Fidelity’s own Q2 2026 Retirement Analysis covers 25.8 million 401(k) participants at Fidelity alone, which puts the millionaire share at roughly 3% even within that already-saving population — before counting the tens of millions of American families who have no workplace retirement account at all.
The Fed’s survey puts a number on that gap: 54.9% of American families held a retirement account of any kind in 2025, up only slightly from 54.3% in 2022. That means 45.1% of families — nearly half — had no 401(k), IRA, or similar account whatsoever. Age does most of the remaining work. Among families that do hold an account, the Fed’s data shows the average balance climbing steeply with age: $48,400 for households headed by someone under 35, $182,400 for 35-to-44, $415,800 for 45-to-54, and $670,200 for 55-to-64 — a nearly 14-fold gap between the youngest and oldest brackets the Fed breaks out. None of that is one age group being smarter than another; it’s mostly decades of compounding and rising income doing arithmetic that looks dramatic once you let it run long enough.
Why millions of workers keep funneling money in
The appeal isn’t complicated, even if the balances end up looking impressive. Contributions to a traditional 401(k) or IRA come out pretax, which lowers your taxable income now in exchange for paying tax on withdrawals later. Most employers that offer a match are, in effect, handing out free money for showing up: Fidelity reports that 81.2% of 401(k) savers are contributing enough to capture their employer’s full match. And since the money is deducted from a paycheck before it ever reaches a checking account, saving happens automatically rather than requiring a monthly decision — reinforced further by auto-enrollment and auto-escalation features that now default millions of new hires into saving without ever filling out a form.
That inertia cuts both ways, though. Vanguard’s “How America Saves” research found that hardship withdrawals — people pulling money out of retirement accounts early, typically to avoid foreclosure or cover medical bills — rose from 1.7% of participants in 2020 to a record 6% in 2025. The same payroll-deduction system that quietly builds seven-figure balances at the top is, for a meaningful slice of savers, also the emergency fund of last resort.
How the account actually grows — and why 55 looks nothing like 25
The jump between age brackets isn’t mysterious once you separate the two things driving it: more money going in, and more time for what’s already in there to grow. A worker in their 20s or early 30s is typically contributing a smaller dollar amount (lower salary, shorter tenure, sometimes still paying down debt or saving for a house) against a small account balance, so market returns on that balance are small in dollar terms even when the percentage return is identical to everyone else’s. By a worker’s 50s, the contribution base is usually larger, the account balance itself is much larger, and every year of average market performance adds a much bigger absolute dollar amount than it did two decades earlier — the same compounding principle that makes early contributions disproportionately valuable in hindsight, even though they rarely feel significant at the time.
That’s the mechanical reason the Fed’s four age brackets roughly double, then nearly double again: $48,400 to $182,400 is a 3.8x jump from the under-35 bracket to the 35-44 bracket; $182,400 to $415,800 is another 2.3x; $415,800 to $670,200 is another 1.6x. The percentage jump shrinks each decade even as the dollar jump stays large, because a bigger base needs a smaller percentage gain to add the same number of dollars.
What’s actually saved, bracket by bracket
Putting the two data systems side by side is the most useful way to read this. The Fed’s 2025 survey, among families that hold a retirement account, reports average balances of $48,400 (under 35), $182,400 (35–44), $415,800 (45–54), and $670,200 (55–64) — the Fed doesn’t break out a 65-and-older row in this particular table, since that section of the bulletin is limited to households headed by someone 64 or younger. Across all ages combined, the conditional median — the middle value, half above and half below — was $106,000 in 2025, up from $95,200 in 2022. The conditional mean for the same group, by contrast, was $451,100, up from $365,700 in 2022 — more than four times the median, and the gap that makes “average” such a misleading word here.
Fidelity’s point-in-time provider data, covering its own much larger but self-selected base of active savers, put the average 401(k) balance at $155,800 as of Q2 2026 (up 10% from the prior quarter and 13% from a year earlier), the average IRA at $144,523 (up 10% both quarter-over-quarter and year-over-year), and the average 403(b) at $145,000. Fidelity also reports that women who have held a 401(k) continuously for at least five years carry an average balance of $273,400 — well above the firm’s blended average, a reminder that tenure matters as much as any other single variable in this data.
One age bracket moved in the wrong direction. The under-35 average fell from $53,800 in 2022 to $48,400 in 2025 — a roughly 10% decline, the only age group in the Fed’s table that didn’t rise. The bulletin doesn’t spell out a specific cause for that bracket alone, but it’s consistent with a cohort effect: every survey year, a fresh crop of 22- and 23-year-olds enters the under-35 bracket with a near-zero balance, pulling the group’s average down even as individuals who age out of it keep growing their own accounts.
Hidden costs and tradeoffs behind the averages
The biggest hidden cost in this data is reading it wrong — treating a provider’s mean as “normal,” when by definition a mean among only the people who already have accounts at a major brokerage excludes the 45.1% of American families with no retirement account at all. A second is timing risk: a meaningful share of this quarter’s 19% jump in millionaire accounts rode a 15% quarterly gain in the S&P 500, which means the same math runs in reverse in a down quarter — the 401(k)-millionaire count has dropped before, falling from a prior high before rebuilding to this one.
A third cost shows up at the other end of the balance sheet. Vanguard’s finding that hardship withdrawals hit a record 6% of participants in 2025, up from 1.7% in 2020, is the flip side of the compounding story above: for a growing share of savers, the account that’s supposed to be left alone for thirty years is being tapped years or decades early, which resets the compounding clock exactly when it has the least time left to recover.
What people get wrong about “the average retirement account”
Treating a provider’s mean as the typical balance. Fidelity and Vanguard’s averages describe people who already have an account with them — a population that, by definition, already saves. They say nothing about the 45.1% of American families the Fed finds hold no retirement account at all, and folding that group in would pull every provider average down substantially.
Confusing “retirement plan participation” with “holding a retirement account.” The Fed’s survey tracks two related but different figures: 65.1% of families in 2025 had access to or participated in a retirement plan of any kind, including old-style defined-benefit pensions, while only 54.9% actually held a 401(k)/IRA-type account. The gap between those two numbers is mostly pension coverage, which pays a monthly benefit rather than accumulating a balance you can check online — a different kind of retirement security that doesn’t show up in any “average balance” figure at all.
Treating Fidelity’s savings-factor guideline as a pass/fail test. Fidelity’s own retirement planning guidance suggests aiming for roughly 1x your salary saved by 30, 3x by 40, 6x by 50, 8x by 60, and 10x by 67 — but that’s built on a specific, stated scenario: saving 15% of income starting at 25 (including any match), keeping more than half your portfolio in stocks over your lifetime, retiring at 67, and targeting a 45% income-replacement rate funded by a 4%-to-5% annual withdrawal. Fidelity itself calls the milestones “aspirational,” not a minimum requirement — change any one of those assumptions (a later start, an earlier retirement, a different spending target) and the “right” multiple for your age changes with it.
Treating 401(k) millionaires as the typical saver. At roughly 3% of even Fidelity’s own actively-saving participant base, the 769,000 figure describes the far right tail of a long distribution, not the center of it. The center, per the Fed’s own national survey, is a family with a retirement account worth $106,000 — a meaningful sum, and nowhere near seven figures.
Bottom line
The Million Dollar Question, answered: C. Fidelity counted a record 769,000 401(k) accounts worth $1 million or more as of mid-2026, up 19% in a single quarter — the fastest quarterly gain since late 2023, boosted heavily by a strong run in the stock market. That number is real, and it’s growing. It’s also the tail, not the center. For the roughly 55% of American families who hold a retirement account at all, the Federal Reserve’s own 2025 survey puts the typical (median) balance at $106,000, and even the age bracket with the highest reported average — 55-to-64-year-olds — sits at $670,200, a long way from a million. Retirement wealth in America isn’t a single line every age group climbs together; it’s closer to a widening fork, with a fast-growing population compounding toward seven figures on one side, and a much larger population — younger savers especially — whose balances are flat, falling, or simply don’t exist yet, even as the headlines keep reporting records.
Related reading: Average Net Worth by Age — and Where Millionaires Sit on the Curve · The Accidental Millionaires: What TSP, 401(k), and Payroll-Deduction Wealth Actually Look Like · Liquidity: How Much Cash the Wealthy Actually Keep · Is $1 Million Enough to Retire? The 4% Rule, Run Honestly · HENRY: $500K and Still Paycheck-to-Paycheck
Methods and sources. The 769,000 401(k)-millionaire count, the 19% quarterly increase, the Q2 2026 S&P 500 performance figure, and the characterization of it as the largest quarterly jump since late 2023 are from Bloomberg News’ reporting via Bloomberg Tax (Sarah Foster and Michelle Amponsah, published September 3, 2026). Average 401(k), IRA, and 403(b) balances, the participant count, the employer-match statistic, and the five-year-tenure balance for women are from Fidelity Investments’ own Q2 2026 Retirement Analysis. The Fidelity savings-factor guideline (1x by 30 through 10x by 67) and its stated assumptions are from Fidelity Viewpoints, “How much do I need to retire?” (published September 25, 2026). The retirement-account holding rate (54.9% in 2025, 54.3% in 2022), the retirement-plan-participation rate (65.1%), the conditional median and mean balances ($106,000 and $451,100 in 2025, versus $95,200 and $365,700 in 2022), and the by-age mean balances among account holders are from the Federal Reserve Board’s Aladangady, Bricker, Chang, Goodman, Li, Moore, Reber, Volz, and Windle, “Changes in U.S. Family Finances from 2022 to 2025” (October 2026). The hardship-withdrawal rate (1.7% in 2020 to a record 6% in 2025) is from Vanguard’s “How America Saves 2026” report as summarized by the National Reverse Mortgage Lenders Association. This is not financial advice; retirement savings benchmarks vary enormously by income, retirement age, and personal circumstance, and anyone making a financial decision based on these figures should consult a licensed financial advisor. See Editorial Standards.
