Roth conversion and IRMAA

A Roth conversion you do in 2026 will show up in your 2028 Medicare premiums. Not 2027. Not the same year. Two years later. This two-year lookback is the detail that breaks naive conversion math, and it is the reason that a conversion planned entirely on this year's bracket can produce a cost that does not appear until the third calendar year after you do it.

The full threshold and surcharge tables for all filing statuses are on the IRMAA brackets page.

How IRMAA works: the mechanism

The Income-Related Monthly Adjustment Amount (IRMAA) is a surcharge on Medicare Part B and Part D premiums assessed on beneficiaries whose Modified Adjusted Gross Income (MAGI) exceeds certain thresholds. The Social Security Administration sets the surcharge for year N using the tax return filed for year N-2. This means 2026 MAGI, as reported on the 2026 Form 1040 filed in April 2027, determines 2028 Medicare premiums. A conversion that lifts 2026 MAGI above a tier boundary generates no premium increase in 2026 or 2027. The full cost appears in 2028, and it runs for as long as MAGI stays in that tier.

The 2026 MFJ IRMAA thresholds and monthly premiums per enrollee, from the CMS 2026 Medicare fact sheet:

The tier boundaries shown above apply to 2026 Medicare premiums, which are based on 2024 MAGI. If you are planning 2026 conversions and want to know the impact on 2028 premiums, the relevant question is whether 2026 MAGI crosses any tier. The thresholds for 2028 premiums will be indexed from 2026 levels, but for planning purposes the current $218,000 MFJ boundary is the right reference until updated figures are published.

Worked example: the one-year cliff

A married couple, one spouse born 1961 (age 65 in 2026, enrolling in Medicare this year), one spouse born 1962 (age 64, enrolling in 2027). Pension income of $180,000 per year provides their living expenses. Their 2026 MAGI before any Roth conversion is $180,000, which is comfortably below the $218,000 MFJ tier 1 boundary. They hold $1,200,000 in a traditional IRA and are considering a $50,000 conversion.

Adding the $50,000 conversion makes 2026 MAGI $230,000, which crosses the $218,000 MFJ threshold. The conversion itself is taxed in 2026 at the 22% marginal rate (their top rate at that income level). The IRMAA consequence shows up differently for each spouse:

The age-65 spouse (born 1961) enrolled in Medicare in 2026. Their 2026 premiums are based on 2024 MAGI, so the 2026 conversion does not affect their 2026 or 2027 premiums. Their 2028 premiums, however, are set by 2026 MAGI: $230,000 crosses the tier 1 threshold. The 2028 Part B premium rises from $202.90 to $284.10 per month, and Part D rises from $0 to $14.50 per month. The annual surcharge for that enrollee is $1,148.40 in 2028 (twelve months at $81.20 for Part B, plus $14.50 for Part D). Running this through the calculator:

The 2026 conversion cost $50,000 at 22% in federal tax, or $11,000 in incremental tax above what this income would have cost at the 22% top of the previous bracket slice. The IRMAA consequence is an additional $1,148.40 in 2028 for the enrolled spouse. If the following year (2027) also produces MAGI above $218,000 for the second spouse's enrollment year lookback, the surcharge compounds. In this example, the 2027 conversion was also $50,000, pushing 2027 MAGI to $230,000, which then determines 2029 premiums for both spouses. The calculator captures this year-by-year as IRMAA crossings.

The asymmetry that matters

The 2-year lookback is asymmetric in an important way. If you convert in a year when you have not yet enrolled in Medicare, the premium consequence arrives two years later when (if) you have enrolled. A filer who does a large conversion at age 63 and enrolls at 65 will see the IRMAA surcharge in their first year on Medicare, which coincides with adjusting to the premium system for the first time. The surcharge is not retroactive to prior years of Medicare enrollment; it applies only to years where both enrollment and the lookback are active. This is why the age at which you convert matters almost as much as the amount.

It also means the tier boundary is not a symmetric cost on both sides. A conversion that takes MAGI from $215,000 to $220,000 crosses the first MFJ tier ($218,000) for a total conversion increment of $5,000 but produces $1,148.40 per enrolled spouse per year in additional premium. That is, for a couple with both spouses on Medicare, the marginal IRMAA cost of the last $2,000 of conversion above the tier is $2,296.80 per year for as long as MAGI stays above the boundary. A single $2,000 overshoot held for three years costs $6,890 in IRMAA alone.

IRMAA Life-Change Exception: when you can appeal

SSA Form SSA-44 allows a beneficiary to request that IRMAA be assessed on more recent income if a qualifying life-changing event (retirement, death of spouse, divorce, reduction in work hours) caused income to drop significantly in the two years since the base year. A Roth conversion does not qualify as a life-changing event; conversions are discretionary. The SSA-44 process exists for income that fell, not income that was deliberately elevated. This is worth knowing because readers sometimes assume that explaining the conversion to SSA will result in a waiver. It will not.

The smoothing trap

A common mistake is trying to stay just under the IRMAA tier boundary each year to avoid surcharges entirely. For a couple with a $1.2M traditional IRA, a pension of $180,000, and a ten-year conversion window before the RBD, staying under $218,000 limits annual conversions to roughly $38,000 ($218,000 minus $180,000 pension). At that pace, the couple converts approximately $380,000 over ten years. On a $1.2M account growing at 6% annually, the unconverted balance ten years out is roughly $1.7M, and the RMDs that begin at age 75 would push ordinary income well above $218,000 anyway. In that situation, the lifetime IRMAA cost of the smoothing strategy is often higher than accepting tier 1 IRMAA during a concentrated conversion window and shrinking the account aggressively. The calculator's scenario comparison tests this directly: run fill-22 (which ignores tier boundaries) against a custom-low conversion and compare the combined lifetime tax and IRMAA. The result is situation-specific, but for high-balance accounts the aggressive conversion usually wins even accounting for the tier surcharges.

Related reading

How the annual conversion amount changes as Social Security comes online and RMDs approach is covered on How much should I convert to Roth this year? The SS provisional income interaction that raises the effective marginal rate above the nominal bracket is covered on Roth conversion and Social Security. For the foundational mechanics of bracket filling, What is bracket filling? has the full worked example with a default MFJ scenario. How the IRMAA exposure window varies by birth cohort depending on the SECURE 2.0 RBD is covered in The Roth conversion window.

Run your own numbers.