How Much Should You Save for Retirement?

There is no single correct retirement number. A useful U.S. estimate starts with how much you may spend each year in retirement, how long that spending might need to last, and what you can save between now and then. Investment returns are uncertain, so any target should be treated as a planning range. Rules of thumb such as “$1 million” or “15% of pay” are conversation starters, not personalized answers. This guide does not promise that a calculator result will be reached or that a withdrawal rule will last for every market path.

Start with spending, not a round number

Some people use a replacement-rate shortcut, such as planning for 70% to 80% of current gross income. That can overshoot or undershoot. A tighter method is to sketch retirement spending in today’s dollars: housing, property tax and insurance, health care including Medicare premiums, food, transportation, travel, and taxes on withdrawals.

Spending is not a flat line. Early years may include travel; later years may include more health costs. You need a number that would still feel tight if inflation or medical bills ran hotter than you hoped. Do not subtract a mortgage or assume a downsize unless those changes would actually happen.

Today’s dollars versus future dollars

Write the spending target in today’s dollars first. Then either inflate it to the retirement year at an assumed inflation rate, or keep everything in today’s dollars and use a real (inflation-adjusted) return in the growth engine. Mixing a nominal 7% return with a spending figure that never inflates will overstate how long the money lasts. Pick one convention and stick with it in a given scenario.

Social Security, pensions, and the annual gap

After you have a spending sketch, subtract income you consider relatively stable. For many households that is a cautious Social Security estimate, a defined-benefit pension if you have one, or a portion of a spouse’s benefit. The remainder is the annual gap that invested savings, part-time work, or a paid-off house (by cutting housing cost) may need to support.

Social Security is not a bond and not a stock. Claiming age changes the monthly amount. Working longer can raise the benefit and shorten the years you must fund. Pensions vary: some adjust for inflation, many do not, and survivor options can reduce the check. If you have no pension, invested savings are doing more of the job, which usually means a larger target or more flexible spending.

Turn annual needs into a savings target

A simple planning rule is to multiply the annual gap by 25, which relates to a 4% first-year withdrawal. That guideline comes from historical U.S. market research on a particular mix and period. It is not a guarantee that 4% will last 30 years for you. A more conservative household might use 30 times the annual gap.

If the annual gap is $40,000 in today’s dollars, 25× is $1,000,000 and 30× is $1,200,000. The multiple does not tell you how to invest. It translates a spending gap into a ballpark so you can test contributions.

Taxes matter. Traditional 401(k) and IRA withdrawals are generally taxable as ordinary income. Qualified Roth withdrawals are generally tax-free. A $40,000 spending gap funded from a pre-tax account may require a larger gross withdrawal. Required minimum distributions can force extra taxable income later even if you did not want to spend it.

Work backward to a monthly contribution

Once you have a target, the retirement calculator can estimate whether current savings plus monthly contributions may get there under a constant return assumption. If the projection falls short, the levers are saving more, working longer, spending less in retirement, or taking more investment risk—each with tradeoffs and none with a promised outcome.

A practical sequence:

1. Estimate annual retirement spending in today’s dollars.
2. Subtract expected Social Security or pension income you are willing to count, using conservative figures.
3. Convert the remaining need into a savings target using a conservative multiple such as 25× or 30×.
4. Test contributions and timelines with more than one assumed return, including a lower one, and with more than one retirement age.

Employer matching contributions are part of the saving rate when you actually receive them. If you contribute 6% and the employer adds 3%, the workplace rate is 9% of pay into that plan, not 6%. Match formulas differ. Contribute enough to capture a true match before you worry about fine-tuning IRA versus brokerage, unless high-interest debt or a missing emergency fund is the more urgent hole.

Accounts, match, and competing cash needs

A 401(k), 403(b), or similar workplace plan is often the easiest automatic path. IRAs (Roth or traditional) add a second bucket with their own limits and income rules. Health savings accounts, if you are eligible, can play a retirement-health role because unused balances may be invested and later used for qualified medical costs. None of these accounts guarantees a return. They change tax timing and, in a workplace plan, may add a match.

High-interest credit card balances can consume cash that looks like a retirement contribution on a spreadsheet. If card APR is in the high teens or more, paying that down often improves the plan more than raising an assumed investment return. A larger house can also raise the retirement spending gap through taxes, insurance, and maintenance.

Expected return is an assumption. Sequence of returns, inflation, fees, and taxes can all change the outcome. Use calculator results as a discussion tool, not as a promise that a target will be reached.

Worked example: age 40, retire at 65, $800 a month

This example is a scenario with a constant return. It is not a forecast of the stock market and not a typical result.

Suppose you are 40, hope to retire at 65, have $120,000 saved, and can contribute $800 a month including any employer match you are counting. That is 25 years, or 300 contributions, totaling $240,000 of new deposits plus the $120,000 already saved.

If you assume a 6% average annual return, a retirement calculator will show an estimated balance at 65. Under a simple compound-interest style illustration, that ending figure can land in the neighborhood of $900,000 depending on compounding conventions. That is an order of magnitude, not a promised $900,000.

Running the same inputs at 4% shows a much lower illustration—often closer to the mid-$600,000s in this kind of setup. If the spending gap was $40,000 and you used 25×, the target was $1,000,000, and both scenarios are short. Raising contributions, delaying retirement, or reducing the gap are the levers. None of them is guaranteed to close the gap in real markets. Label every rate.

What to do if the projection falls short

Shortfalls are information. They are not a verdict that you “failed” at a round number. Common adjustments:

• Increase the automatic contribution when you get a raise, even by 1% of pay.
• Capture the full employer match if you were leaving it on the table.
• Delay retirement or plan a gradual step-down in work, which both adds saving years and shortens the drawdown period.
• Reduce the retirement spending target in ways you would actually accept, such as a cheaper location or a paid-off house.
• Review fees in the 401(k) fund list; a lower-cost fund does not raise expected return by a promised amount, but costs are certain.

Taking more stock-market risk can raise the assumed return in a calculator and also raise the chance of a large decline near retirement. Saving more at a moderate allocation often does more for a shortfall than switching the assumed return from 6% to 8%.

Limits of retirement calculators

Retirement tools on this site and elsewhere assume a constant rate or a simple growth path unless they say otherwise. They typically do not fully model:

• Sequence of returns in retirement, when withdrawals can lock in losses.
• Inflation on each spending category at different rates (health care versus groceries).
• Taxes at the federal and state level, including IRMAA Medicare surcharges.
• Longevity: a plan to age 90 is not a promise you will need 30 years or only 20.
• Policy changes to Social Security, RMDs, or contribution limits.
• Job loss, divorce, or a long pause in contributions.

Past U.S. market averages are not a promise of future compound returns. A 4% withdrawal illustration is not insurance. If the plan only works at a high return, a late claiming age, and a low spending figure all at once, you have a fragile plan. Strengthen it by saving more, spending less in the sketch, or working longer—not by typing a friendlier yield. Confirm contribution limits with current plan documents and IRS rules.