How Much of Your Social Security Is Taxable?

Somewhere between none of it and 85 percent of it. Which end of that range you land on is decided by a single measure, and the measure is not your tax bracket, your income tax return’s bottom line, or your Social Security check itself.

That is the short answer. The rest of this page is the measure, the two tiers it feeds, and the handful of things that quietly push a household from one tier to the next.

One clarification first, because it is the single most common misunderstanding on this topic and it changes how the whole page reads.

Is 85 percent a tax rate?

No. It is a ceiling on how much of the benefit can be pulled into your income, not a rate applied to the benefit.

The statute is unusually clear about this. The amount included in gross income is the lesser of a computed sum or 85 percent of the Social Security benefits received during the taxable year (26 U.S.C. §86(a)(2)(B)). Included in gross income. That included amount then sits on your return alongside everything else and is taxed at whatever ordinary rates apply to you.

So a household at the top of this calculation is not paying 85 percent tax on its benefit. It is reporting up to 85 percent of the benefit as income, and paying its own marginal rate on that portion. If that household is in a modest bracket, the actual tax on the benefit is a fraction of the headline number.

The mirror image is worth stating too. Because the ceiling is 85 percent, at least 15 percent of a Social Security benefit stays outside gross income no matter how high the rest of your income goes. There is no version of this where the whole benefit becomes taxable.

What is provisional income?

It is the measuring stick, and it is not a number that appears anywhere on your tax return.

The statute builds it in one sentence. A taxpayer is reached by this section if the sum of modified adjusted gross income for the taxable year, plus one-half of the Social Security benefits received during the taxable year, exceeds the base amount (26 U.S.C. §86(b)(1)). That sum is what practitioners call provisional income. The IRS calls it combined income. They are the same figure.

Two pieces, then.

Modified adjusted gross income. Your adjusted gross income, computed without regard to this section and to a short list of other exclusions, and then increased by tax-exempt interest (26 U.S.C. §86(b)(2)). More on that increase in a moment, because it is where most of the surprise lives.

Half of the benefit. Only half of what Social Security paid you goes into the measure. This is a small mercy and an easy thing to get wrong in your own head. A household that adds the full benefit to its other income will conclude the situation is worse than it is.

Notice what provisional income is not. It is not taxable income. Taxable income sits below your standard deduction. Provisional income starts from adjusted gross income, which sits above it, and then adds things back. The two figures are never the same, and planning aimed at one is not aimed at the other.

Why does municipal bond interest count?

Because the statute says so, in plain words, and this catches a lot of thoughtful people.

Modified adjusted gross income here is adjusted gross income increased by the amount of interest received or accrued during the taxable year which is exempt from tax (26 U.S.C. §86(b)(2)(B)). Tax-exempt interest. Municipal bond income.

Read that again if you own municipal bonds. That interest is invisible to the income tax and completely visible to this measure. A portfolio built specifically to keep reported income low can still be raising the share of the Social Security benefit that lands on the return, and nothing on the return shows it happening.

I am not saying that makes municipal bonds a poor holding. Whether they belong in a portfolio is a different question with its own answer, and it depends on far more than this one interaction. I am saying that the tax-free label describes the income tax and does not describe this calculation, and a household that has never been told that is working from an incomplete picture.

The same logic runs the other way, and it is the more useful half. Money that never enters adjusted gross income never enters provisional income either. A qualified Roth withdrawal is outside it. A return of your own basis is outside it. Those dollars can be spent without moving the measure at all.

How do the two tiers actually work?

There are two of them, and almost every plain-English description of this topic collapses them into one.

Below the base amount, none of the benefit is taxed. The section only reaches a taxpayer whose provisional income exceeds the base amount, and the base amount is set separately for a joint return and for everyone else (26 U.S.C. §86(c)(1)). A household under its own figure includes nothing.

Between the base amount and the adjusted base amount, the included portion is capped at half. In this middle band, the amount included in gross income is the lesser of one-half of the benefits received, or one-half of the excess of provisional income over the base amount (26 U.S.C. §86(a)(1)). Two things follow from that. The included amount starts at zero at the threshold and climbs from there, so crossing the line does not detonate anything. And within this band, half of the benefit is the most that can ever be included.

Above the adjusted base amount, the 85 percent tier begins. Here the amount included in gross income is the lesser of two figures: a sum built from 85 percent of the excess over the adjusted base amount plus a carried-forward piece from the first tier, or 85 percent of the benefits received for the year (26 U.S.C. §86(a)(2)). The adjusted base amount, like the base amount, is set separately for a joint return (26 U.S.C. §86(c)(2)).

The exact dollar figures for both thresholds, at both filing statuses, are written into §86(c) itself, and there are only four of them.

• Base amount — $32,000 on a joint return (26 U.S.C. §86(c)(1)(B)), and $25,000 otherwise (26 U.S.C. §86(c)(1)(A)).
• Adjusted base amount — $44,000 on a joint return (26 U.S.C. §86(c)(2)(B)), and $34,000 otherwise (26 U.S.C. §86(c)(2)(A)).

Those four are the whole threshold set. Below your base amount, none of the benefit is included. Between the two, up to half. Above the adjusted base amount, the upper tier applies.

Here is what that structure means in practice, and it is the part worth carrying. This is a ramp, not a cliff. In the first band the included amount rises with half of each additional dollar of provisional income. In the second it rises with 85 cents of each additional dollar, until the ceiling is reached and it stops rising entirely. Nobody falls off an edge. But the slope in the upper band is steep, and a household sitting in it is effectively adding 85 cents of benefit income to its return for every extra dollar it creates elsewhere.

If it helps to have the numbers in one place, I keep them all in the Ultimate Retirement Guide.

Why don’t the thresholds change every year?

Because Congress wrote fixed dollar amounts into the statute and did not attach an inflation adjustment to them.

That is genuinely unusual. Tax brackets move each year. The standard deduction moves. Medicare surcharge thresholds move. The figures in §86(c) do not. They are written as flat dollar amounts, with no indexing provision anywhere in the section, and they have sat unchanged since the upper tier was added in 1993 (Pub. L. 103-66, §13215).

The consequence compounds quietly. Every year that benefits rise with a cost-of-living adjustment and portfolio income rises with everything else, more households cross a line that has not moved. Nobody legislated that expansion. It happens on its own, by drift, and it is the reason a rule originally aimed at higher-income retirees now reaches a great many ordinary ones.

There is one useful side effect for a reader. Because these thresholds do not move, a page describing them does not go stale the way a bracket table does. What is written in §86(c) today will very likely be the same figure years from now, absent an act of Congress.

What happens if a married couple files separately?

This is the trap, and it is written into the statute in two places.

For a married taxpayer who does not file a joint return and who does not live apart from their spouse at all times during the taxable year, the base amount is zero (26 U.S.C. §86(c)(1)(C)). The adjusted base amount is zero as well (26 U.S.C. §86(c)(2)(C)).

Zero. There is no protected band at all. With no threshold to clear, the calculation begins at the first dollar, and a married-filing-separately household living together will generally see the maximum share of the benefit included.

Note the exact condition, because it is doing real work. The zero applies where the couple did not live apart at all times during the year. A married person who lived apart from their spouse for the entire taxable year and files separately is treated under the ordinary rules instead. That distinction turns on facts, not on a checkbox, and it is one of the places where the general explanation genuinely runs out and a specific situation needs specific attention.

Filing separately is sometimes chosen for reasons that have nothing to do with this section — an income-driven student loan calculation, a liability concern, a separation in progress. The point here is only that this consequence rides along with that choice, and it is not obvious from the return itself.

How does a Roth conversion or an IRA withdrawal fit in?

Both are ordinary income. Ordinary income lands in adjusted gross income. Adjusted gross income is the base of provisional income. So both raise the measure directly, dollar for dollar.

That produces an interaction people often meet by surprise. In the upper band, a dollar of conversion income is not just a taxable dollar. It also pulls up to 85 cents of benefit income onto the return with it. The effective cost of that conversion dollar is therefore higher than the bracket alone suggests, sometimes considerably higher, until the benefit reaches its ceiling and the effect switches off.

This paragraph is illustrative and directional — no dollar amounts, no real household. Picture two identical retirees with the same benefit and the same portfolio. One converts nothing this year. The other converts an amount that carries provisional income from the middle band up into the upper band. The second retiree owes tax on the conversion, which was expected, and also finds that a larger share of the benefit has become part of the return, which usually was not. Neither outcome is a mistake. They are simply two effects of one decision, and a plan that counted only the first has understated the year.

The same arithmetic applies to any ordinary-income event: a large withdrawal from a traditional retirement account, a required minimum distribution once those begin, a pension starting, a realized capital gain adding to adjusted gross income. None of them are special cases in this section. They are income, and income is what the measure reads.

The reverse also holds. Because the ceiling exists, a household already at 85 percent cannot be pushed higher by this section. Once the maximum share of the benefit is included, additional income affects the ordinary tax calculation and stops affecting this one. That is a meaningful planning fact in both directions, and it is why the position of a specific household within the tiers matters more than the general rule.

What do people get wrong about this?

They think 85 percent is the tax rate. It is the maximum share of the benefit that can be included in income (26 U.S.C. §86(a)(2)(B)). The rate that applies to it is their own.

They think the whole benefit can become taxable. It cannot. The ceiling leaves at least 15 percent of the benefit outside gross income permanently.

They add the full benefit to the measure. Only one-half of benefits received goes into provisional income (26 U.S.C. §86(b)(1)(A)).

They think tax-exempt interest is invisible. It is invisible to the income tax and added back in full here (26 U.S.C. §86(b)(2)(B)).

They think crossing a threshold is a cliff. It is a ramp. The included amount starts at zero at the base amount and rises with income from there.

They expect the numbers to move each year. They do not. The figures in §86(c) are not indexed, which is why the share of retirees affected keeps growing without any change in the law.

They treat one MAGI as every MAGI. The modified adjusted gross income defined in this section is specific to it. The measure that decides a Medicare surcharge is a different definition with a different add-back list, on a page of its own.

---

This is a rule about a measure, and the measure is more sensitive to the rest of your income than to your benefit. That is why the taxation of benefits usually shows up as a consequence of some other decision rather than as a decision of its own.

For the claiming side of the question — when to file, what delaying buys, and what break-even analysis does and does not see — start at the break-even page. For everything around it, the Social Security hub has the rest.