Withdrawal Sequencing: A Complete Guide to Tax-Smart Retirement Drawdowns

What withdrawal sequencing is

Retirement typically funds itself from three account types with three different tax treatments:

Taxable brokerage accounts. Principal is already taxed. Only gains (capital gains on sale, dividends, interest) are taxable events. Most growth benefits from preferential long-term capital gains rates rather than ordinary income rates.

Traditional tax-deferred accounts (traditional IRA, 401(k), 403(b), 457). Contributions were pre-tax. All withdrawals are ordinary income at your marginal rate. Required Minimum Distributions (RMDs) start at age 73 or 75 under SECURE 2.0.

Roth accounts (Roth IRA, Roth 401(k)). Contributions were post-tax. Qualified withdrawals are tax-free. No RMDs during the original owner’s lifetime (Roth 401(k) RMDs were eliminated starting 2024 under SECURE 2.0).

Withdrawal sequencing is the question of which account to withdraw from each year, in what proportion. The choices compound across 20–30 years of retirement. For a household with meaningful balances across all three account types, the difference between good and poor sequencing can run into six-figure lifetime tax savings.

The conventional default

The textbook retirement withdrawal order, repeated across most retail planning resources:

1. Taxable first — spend down the already-taxed accounts
2. Traditional second — tap tax-deferred accounts, paying ordinary income
3. Roth last — let tax-free growth compound longest

The logic: tax deferral compounds. Money growing inside a tax-advantaged account for an extra 10 years beats the same money in a taxable account where gains and dividends get taxed annually. Delaying withdrawals from Roth and traditional maximizes the tax-sheltered compounding period.

This isn’t wrong — it just answers a narrow version of the question. It optimizes for “how do I delay taxes as long as possible,” which is often not what you actually want to optimize.

Why the default is often wrong

The conventional sequence treats each year as independent: which account has the lowest marginal tax cost right now? That framing misses the compounding effects of the years you’re not thinking about.

Problem 1: RMD stacking. If you defer all traditional withdrawals until RMDs force your hand, the RMDs at 73+ may be very large — because the account kept growing for the intervening decade. Those RMDs arrive on top of Social Security, pension income, dividends, and anything else taxable. The resulting combined income often pushes retirees into the 24% or 32% bracket and triggers IRMAA tier jumps — years after they could have paid tax on the same dollars at 12% or less.

Problem 2: Lost bracket arbitrage. For most retirees, the years between retirement and Social Security (and especially before RMDs) are the lowest-taxable-income years of their adult life. The 12% and 22% brackets are enormous — in 2026, the 12% bracket runs up to $100,800 for married filers. A retiree with only portfolio income and no Social Security can intentionally pull from traditional accounts in this window at rates far below what they’ll face later.

Problem 3: Survivor penalty ignored. When one spouse dies, the surviving spouse typically moves from MFJ to single filing status. Brackets compress dramatically — the 22% bracket for single filers in 2026 tops out at $105,700, compared to $211,400 for MFJ. Income that was comfortably in the 22% bracket as a couple may be in the 32% or 35% bracket for a widow(er). Deferring all traditional distributions increases the surviving spouse’s lifetime tax burden.

Problem 4: Legacy concentration. A retiree who never touches traditional IRA money dies with a large pre-tax balance that heirs must distribute within 10 years under SECURE Act rules — at the heirs’ marginal rates during their own peak earning years. Deferral can push a tax bill onto the next generation at worse rates than the retiree would have paid.

The smarter default

For most retirees with meaningful traditional IRA balances ($500K+), a better baseline:

Pre-retirement to Social Security claiming age (typically 62–67 or 62–70): Draw primarily from taxable accounts for living expenses, but intentionally add traditional IRA withdrawals up to a tax-bracket ceiling — usually the top of the 12% bracket for MFJ, which is a remarkably large amount of taxable income in 2026. Every dollar pulled at 12% now is a dollar that won’t later be pulled at 22%+ during RMD years.

Social Security claiming age to RMD age (typically 67–73 or 70–75): Continue the same approach but adjust for the new Social Security income in the calculation. Now 85% of Social Security is counted toward combined income. Traditional withdrawals still fill remaining bracket room, but there’s less of it.

RMD years (73+/75+): RMDs take precedence. Traditional withdrawals are no longer elective for the minimum required amount — but you can still strategically add more traditional withdrawals if it avoids bracket jumps in future years. Roth is now available to cap annual taxable income in high-income years.

Throughout: Roth serves as a strategic hedge. Rather than preserving it for last, use it surgically in years when an additional traditional withdrawal would cost more than the nominal marginal rate — because it would push you into IRMAA tier, cross the ACA cliff, or trigger Social Security taxation jumps.

The 2026 tax environment

Two OBBBA changes shape the 2026 withdrawal sequencing environment:

Permanent TCJA brackets. The OBBBA (signed July 4, 2025) made the TCJA tax bracket structure permanent. The 10/12/22/24/32/35/37% brackets are no longer scheduled to sunset. This removes the “convert and withdraw before rates go up” urgency that was driving planning from 2023–2025. Rates are stable.

Temporary senior deduction. Taxpayers age 65+ get an additional $6,000 deduction (single) or $12,000 (MFJ with both 65+), phasing out above $75K/$150K MAGI. This effectively widens the 0% bracket for seniors with moderate income. But the phaseout creates a hidden effective marginal rate spike between $75K–$175K (single) or $150K–$350K (MFJ) MAGI — withdrawals in that range carry the nominal rate plus the cost of losing deduction value. The deduction is scheduled to expire after 2028 under current law.

2026 bracket thresholds (MFJ):

• 10%: up to $24,800
• 12%: up to $100,800
• 22%: up to $211,400
• 24%: up to $403,550
• 32%: up to $512,450

Combined with the standard deduction ($32,200 MFJ in 2026, plus $1,650 per spouse 65+, plus the $12,000 senior deduction if under MAGI phaseout), a 2026 retired couple can have gross income of ~$150K and still be entirely in the 12% bracket or below.

Worked example: Low-tax-bracket filling

Household. Married couple, both 64. Retired. $1.2M traditional IRA, $400K taxable brokerage, $250K Roth IRA. Annual spending need: $100K.

Conventional approach (taxable-first):

• Years 1-4 (age 64-67): Spend down taxable brokerage. Realize only minimal capital gains. AGI ~$5K per year (dividends). Federal tax: near zero.
• Year 5 (age 68): Begin Social Security at $48K combined; start traditional IRA withdrawals to fund spending. AGI jumps to $80-100K.
• Year 10 (age 73): RMDs begin. Initial RMD on the still-large traditional balance: ~$45K. Combined with Social Security and traditional IRA supplemental withdrawals, AGI exceeds $130K annually. Higher brackets, IRMAA risk, survivor penalty looming.

Smart sequencing approach:

• Years 1-4 (age 64-67): Spend taxable brokerage for $60K of annual needs, but also add $45K of traditional IRA withdrawals each year. AGI: ~$50K (after standard deduction). Tax rate: fully in 12% bracket. After standard deduction and senior deduction (once they hit 65), federal tax liability: minimal.
• Year 5-9 (age 68-72): Continue traditional withdrawals in parallel with Social Security. AGI maintained around $100K — still in 12% MFJ bracket.
• Year 10+ (age 73+): RMDs now based on a traditional balance that’s $300-400K smaller than it would have been. RMDs are more manageable, bracket management remains feasible, IRMAA stays out of reach.

Over 30 years of retirement, the smart sequencing approach typically reduces total federal taxes by $50K–$150K depending on specifics. Not a small number.

When the conventional default is actually right

For some households, taxable-first-then-traditional-then-Roth is correct:

Small traditional balance. If the traditional IRA is under $200K, RMDs will never be large enough to create problems. Conventional sequencing is fine.

Very high spending. If the household is spending $250K+ per year and the traditional balance will be fully drawn down within 15-20 years regardless of sequence, optimization between the account types matters less.

Strong bequest motive. If the goal is to leave Roth assets to heirs (Roth IRA is often the best inheritance — 10-year rule applies but distributions are tax-free), preserving Roth makes sense. The conventional sequence does this naturally.

No traditional account. If the household already did extensive Roth conversions pre-retirement or never had significant traditional savings, there’s no traditional balance to optimize around.

The coordination dimension

Withdrawal sequencing doesn’t exist in isolation. It interacts with:

Social Security claiming. Delaying Social Security to 70 creates a longer low-income window for traditional withdrawals or Roth conversions. See Social Security Claiming Strategy.

Roth conversions. The same low-income window that’s good for filling brackets via withdrawals is also good for converting traditional to Roth. Withdrawal and conversion decisions are coordinated. See Roth Conversions: The Complete Guide.

RMD planning. Smart sequencing reduces future RMDs. Pre-RMD traditional withdrawals and Roth conversions both shrink the balance that must later be mandatorily distributed. See RMD Planning: The Complete Guide.

ACA subsidies. Pre-Medicare retirees face the 400% FPL cliff. Traditional withdrawals raise MAGI directly; Roth withdrawals don’t. Sequencing decisions during the 62-to-65 pre-Medicare window interact heavily with ACA subsidy math. See The 400% FPL Cliff.

IRMAA. Medicare premium surcharges use a two-year lookback on MAGI. Traditional withdrawals that push you past an IRMAA tier cost you not just the additional tax but the premium surcharge for two years forward. See IRMAA Tiers and Roth Conversions.

Social Security taxation. The 85% taxation threshold interacts with every withdrawal decision. Each additional dollar of traditional withdrawal may cost more than the nominal rate if it drags more Social Security into taxable status. See Social Security Taxation.

These aren’t independent problems. A competent withdrawal sequence has to consider all of them simultaneously, usually with tools that can model the interactions. Spreadsheet-level analysis breaks down above a handful of variables.

The annual decision framework

A simple year-by-year process for most retirees:

Step 1: Project baseline income for the year. Sum of expected Social Security, pension, RMDs (if applicable), dividends, interest, and any other “required” income.

Step 2: Determine spending need gap. Annual spending minus after-tax baseline income = gap to fund from elective withdrawals.

Step 3: Identify bracket room. How much taxable income can you add before hitting a bracket you want to avoid (next marginal rate, next IRMAA tier, 85% SS taxation threshold if applicable)?

Step 4: Match elective withdrawals to bracket room. If bracket room exceeds gap, fill the gap with traditional IRA withdrawals. If bracket room is less than gap, take traditional up to the ceiling and cover the remainder from Roth or taxable.

Step 5: Consider conversion room. If bracket room exceeds both the gap and any planned traditional withdrawals, add Roth conversion of the remainder — creating future tax-free dollars at today’s low rate.

Step 6: Adjust for specific factors. ACA cliff if pre-Medicare. IRMAA tiers if 63+. Charitable intent (QCDs once 70½). Year-specific income events (large cap gains, property sale).

Step 7: Execute and document. Take the planned withdrawals. Record the rationale. Re-run the framework next year with updated numbers.

This process doesn’t require a PhD in tax optimization. It requires discipline to run it annually and willingness to vary from conventional rules when the math favors it.

Common mistakes

Taking more from traditional than bracket room allows, to “round up” to a convenient figure. The extra $5,000 over the 12% ceiling costs 22% (plus possibly more via cascading effects). Stop at the ceiling.

Treating Roth as sacred. Roth isn’t for worshipping — it’s a tax-rate hedge. Using it surgically in high-tax years is exactly what it’s designed for.

Ignoring state taxes. Some states have income tax brackets that differ sharply from federal. A withdrawal strategy optimized for federal may miss state-level optimization. Critical in Oregon, California, New York, Minnesota, and similar high-tax states.

Forgetting HSA. For retirees with HSA balances and qualified medical expenses (including Medicare premiums post-65), HSA withdrawals for those expenses are tax-free and don’t count as income. Often the most tax-efficient source of funds for medical spending.

Not coordinating between spouses. Each spouse’s IRA is their own. Coordinated timing across two accounts can produce better results than independent optimization.

Late-year surprises. A December realization that your planning assumption was off can leave you with tax consequences you can’t fix until the following year. Monthly or quarterly check-ins prevent December regrets.

Try Ordo

Frequently asked questions

Sources

Chris Gammill is the founder of Ignis Tools and writes about tax-aware retirement planning. Research and drafting assisted by AI tools; all figures and claims verified by the author against primary sources.

1. Fidelity Viewpoints — Tax-savvy withdrawals in retirement — retrieved 2026-04-21
2. Morningstar — Retirement Withdrawal Sequencing Rules of the Road — retrieved 2026-04-21
3. IRS Revenue Procedure 2025-32 — 2026 inflation adjustments (brackets, standard deduction) — retrieved 2026-04-21
4. Nationwide Financial — 2026 Tax Strategies for Retirees & Workers — retrieved 2026-04-21