Which Retirement Account Should You Drain First? Conventional Order Saves Real Tax Dollars
$200K taxable, $600K traditional, $150K Roth, $60,000/year spending need over 25 years — drawing taxable first, then traditional, then Roth last beats a proportional draw-from-everything approach by letting tax-advantaged accounts compound longer before they're tapped.
The order you draw from matters as much as how much you draw
We ran a common three-bucket retirement ($200K taxable, $600K traditional, $150K Roth) against a $60,000/year spending need over 25 years. Drawing taxable first, then traditional, then Roth last — the "conventional" order most fee-only planners default to — beat a proportional draw-from-everything approach on total lifetime tax paid. The reason isn't about which bucket gets taxed at what rate in isolation; it's that leaving tax-advantaged accounts untouched longer lets them compound longer before any of that growth gets taxed (or, for Roth, ever taxed at all).
How the math works
Both strategies fund the same after-tax annual spending need from a pool of three buckets, each grossed up by that bucket's effective tax rate: to net $10,000 after a 15% tax, you'd need to withdraw $10,000 ÷ (1 − 0.15) ≈ $11,765 gross. "Conventional" drains taxable completely before touching traditional, then traditional before touching Roth. "Proportional" draws from all three every year, weighted by each bucket's current balance share.
What this tool doesn't model: progressive tax brackets (a flat effective rate per bucket is a simplification — pulling a huge amount from traditional in one year can push you into a higher bracket than this model assumes), Required Minimum Distributions forcing traditional withdrawals starting at 73 regardless of strategy, Social Security and pension income reducing how much you need to draw from these buckets at all, and state taxes.
Math runs locally. Inputs never leave your browser.Source on github.
Where "conventional always wins" breaks down
- RMDs force traditional withdrawals starting at 73.You can't actually save traditional for last indefinitely — required minimum distributions kick in regardless of your preferred order. A more complete plan blends conventional sequencing with RMD-aware traditional withdrawals in your 70s. See the RMD Calculator for that piece.
- A big traditional withdrawal can jump you into a higher bracket.This tool's flat-rate assumption doesn't capture bracket-filling strategies — some years it's better to draw more from traditional while you're in a low-income year (e.g., before Social Security starts) and fill up to the top of a bracket rather than strictly minimizing that year's withdrawal.
- Estate planning can flip the priority.If leaving money to heirs matters, Roth accounts (no lifetime RMDs, tax-free to heirs within the 10-year rule) are often the accounts you want to preserve longest — which conventional order already does, but for a different reason than tax minimization during your lifetime.
- Your taxable account's effective rate is doing a lot of work in this model.If your taxable account is mostly unrealized gains (low basis), the blended rate is close to your capital-gains rate; if it's mostly cash/bonds already taxed annually, the effective rate on withdrawal is closer to 0%. Get this number right for your actual holdings.
What to actually do with this number
- Run the calculator with your actual account balances and a realistic blended taxable-account rate.
- Treat the conventional order as a strong default, but plan to blend in RMD-driven traditional withdrawals starting at 73 rather than assuming you can defer traditional indefinitely.
- In low-income years (early retirement, before Social Security), consider intentionally drawing more from traditional to fill up a low tax bracket — a refinement this simplified model doesn't optimize for automatically.
- If leaving an inheritance matters, weight preserving Roth balances even more heavily than the tax-minimization math alone suggests.
- Revisit the plan every few years as balances, tax law, and your spending need change.