RSU tax deep dive: what your HR portal won't tell you.

Why your RSU withholding is wrong by default.

When your RSUs vest, the fair market value of the shares on that day becomes ordinary income to you under IRC Section 83. The full vest-day value flows through your W-2 in Box 1 alongside your salary, and your employer withholds federal income tax, FICA, Medicare, and state income tax against that amount before depositing the remaining shares in your brokerage. Mechanically, the vest is treated like a cash bonus that happens to be paid in stock.

The problem is not the mechanism but the rate at which the federal piece gets withheld. Under IRS Publication 15-T, an employer paying supplemental wages — bonuses, commissions, RSU income — has two withholding rates available. Below $1 million of cumulative supplemental wages in a calendar year, the optional flat rate is 22 percent. Above $1 million, the mandatory flat rate is 37 percent. The vast majority of public-company payroll systems default to the 22 percent rate for every employee, because that rate covers the median worker correctly and per-employee tuning of withholding would be operationally expensive at scale.

For an engineer whose total compensation places them in a higher federal bracket, the 22 percent default is structurally insufficient. Single filers cross into the 32 percent bracket above roughly $197,300 of taxable income for tax year 2025 under the IRC Section 1 brackets, and into the 37 percent bracket above $626,350. The married-filing-jointly thresholds run higher, but a household with two technology incomes typically clears them inside a single grant cycle. Once you are above the 22 percent line, every dollar of RSU income that runs through supplemental withholding leaves a portion of the actual liability uncollected, and that uncollected portion compounds across the year as additional vests stack on top of the same shortfall.

The shortfall is not a payroll error and your HR team is unlikely to flag it for you. It is the predictable outcome of a withholding rule that was designed to be administratively simple rather than individually accurate. The responsibility for closing the gap before the April 15 filing deadline falls to you, and the size of the gap is large enough — particularly for engineers vesting six-figure tranches at FAANG-tier companies — that ignoring it puts you on the wrong side of the underpayment penalty under IRC Section 6654. The next section walks through the actual dollar consequences on a representative vest.

The $30,000 April surprise, modeled.

The gap between supplemental withholding and actual liability is easiest to understand on a specific household, so consider a Staff Engineer at a public technology company filing as a single taxpayer. Their compensation for the year consists of $250,000 in base salary, an $80,000 cash performance bonus, and $200,000 of RSU income vesting in October. Total wages reach $530,000, which places the household solidly inside the 37 percent federal bracket.

Once you understand the mechanics on this single example, the same calculation extends predictably. A household vesting $400,000 of RSUs in a 37 percent bracket sees a $60,000 federal gap rather than $30,000. A household vesting $200,000 but sitting in the 32 percent bracket sees a $20,000 gap rather than $30,000. The shape of the problem is the same in every case: the 22 percent line and your actual marginal rate are different numbers, and the difference becomes due in April unless you intervene before then.

Three fixes, and how to choose between them.

The federal gap is solvable, and the IRS does not particularly care which mechanism you use to close it as long as the cash arrives on time. What does matter is the timing: the underpayment penalty under IRC Section 6654 is triggered when the cumulative tax paid through withholding and estimated payments falls below the safe harbor at any point during the year, so the right fix depends primarily on when in the year you become aware of the gap. The three options below are listed in roughly the order you would prefer to use them — earliest intervention to latest.

1. Lift your W-4 withholding at the start of the year. If you have visibility into your expected RSU vest schedule by January, the cleanest approach is to use the IRS Tax Withholding Estimator to model the gap and submit a revised Form W-4 to payroll with an additional fixed dollar amount on Step 4(c). The added withholding is spread across every paycheck for the rest of the year, which keeps household cash flow predictable and removes any need for estimated payments. This is the right answer for engineers whose compensation is stable year over year and whose vest schedule is known in advance.
2. Make a Q4 estimated payment. If the vest hits later in the year, or if you only realize the size of the gap when running preliminary numbers in November, the estimated-payment route is the appropriate fallback. File Form 1040-ES against the modeled gap by January 15 of the following year and the Section 6654 safe harbor is satisfied for the year, even though the underlying tax owed does not change. The estimated payment is functionally equivalent to lifted withholding from the IRS's perspective, but the cash leaves your account in a single transaction rather than across the year.
3. Increase a working spouse's W-4 withholding. Federal withholding is treated by the IRS as paid evenly across the year regardless of when it actually occurred, which makes a spouse's W-4 Step 4(c) unusually flexible for late-year RSU vests. An increase applied in November can be treated as if it had been paid since January, which can rescue a household that missed both the W-4 window and the estimated-payment timing on its own. The mechanic is identical to lifting your own W-4, but the timing flexibility is significantly better.

The Section 6654 safe harbor itself is the floor underneath all three approaches. As long as cumulative withholding plus estimated payments reach 110 percent of prior-year total tax (100 percent if your prior-year AGI was below $150,000) by January 15, the underpayment penalty does not apply, and you have until April 15 to settle the actual balance owed. Because the safe harbor is binary — you either qualify or you do not — it is worth running the calculation explicitly before deciding which lever to pull, so the W-4 update or estimated payment is sized correctly the first time.

For households whose RSU income varies significantly year to year, the prior-year safe harbor often turns out to be the easier target than the current-year liability calculation. If last year's tax was $180,000, paying 110 percent of that — $198,000 — through withholding and estimated payments by January 15 satisfies Section 6654 regardless of what this year's income turns out to be. The remainder gets settled at filing without penalty. For a deeper walkthrough of the month-by-month payment calendar and the W-4 mechanics, see our RSU withholding-gap explainer.

Even with perfect withholding, your 1099-B is wrong.

Closing the withholding gap solves the cash-flow side of the RSU tax problem, but it does not solve the accounting side, and the accounting side is where most self-prepared returns lose money to the IRS without anyone noticing. The issue arises at the second taxable event — when you eventually sell the vested shares — and it depends on a structural quirk of how brokers are required to report cost basis for restricted stock.

When your RSUs vest, the fair market value on that day becomes your cost basis in the shares, because that amount has already been reported as ordinary income on your W-2 Box 1 and the federal, FICA, Medicare, and state taxes on it have already been collected. When you later sell the shares, only the difference between the vest-day basis and the sale price should be subject to capital gains treatment, since the vest-day value has already been taxed as ordinary income. This is how the IRS expects the math to work, and it is how you would expect any reasonable return to handle the situation.

The complication is that the broker's Form 1099-B does not necessarily report the basis the way the IRS expects. Under the cost-basis reporting rules that apply to restricted stock, brokers are prohibited from including the ordinary-income component of vest-day value in the basis they report to the IRS, because doing so would amount to the broker speaking on behalf of the W-2 figure that the employer already reported. The result, in practice, is that the basis line on your 1099-B for RSU sales frequently shows up as $0 or as an unadjusted purchase price that excludes the W-2 amount entirely. If you file Schedule D using that figure unchanged, you are reporting your entire sale proceeds as a capital gain — including the vest-day value that already had ordinary income tax collected on it.

The fix is mechanical and the same across every major brokerage. Fidelity Stock Plan Services, E*TRADE, and Schwab Stock Plan Services all issue a separate supplemental statement alongside the 1099-B, and the supplemental statement carries the corrected vest-day basis for every lot of RSUs that you sold during the year. To clean up the return, you file Form 8949 with adjustment code "B" and the corrected basis from the supplemental statement, and then carry the adjusted gain or loss to Schedule D as usual.

The W-2 Box 14 line, often labeled "RSU" or "ESP/RSU," is informational only and identifies the dollar value of vested RSU income that the employer has included in Box 1. It should reconcile with the supplemental statement once the basis adjustment is applied. When the two figures disagree, the supplemental statement is usually the correct one for filing purposes, but it is worth asking payroll for confirmation before submitting the return. The basis trap is the single most common source of overpayment among tech employees who self-prepare, and Form 1040-X allows you to amend up to three years back, so a basis error caught on a 2023 return is still recoverable today.

How sell-to-cover affects the basis math

The sell-to-cover mechanism — which most public companies use to satisfy the supplemental withholding obligation — interacts with the basis calculation in a way that is worth understanding explicitly. When the employer's broker sells a calculated number of shares at vest day to cover the withholding, those shares are sold at vest-day fair market value, and the broker reports them as a separate sale on the 1099-B. The proceeds equal the vest-day price; the basis for those specific shares is also the vest-day price; the resulting capital gain or loss is approximately zero. This is correct and does not require adjustment.

The basis problem arises on the remaining shares — the ones that actually settle into your account and that you may eventually sell at a different price. Those shares are also basis-adjusted to vest-day fair market value, but the 1099-B will not reflect that adjustment without the supplemental statement, for the same structural reason described above. If you sell those remaining shares in any future year, the same Form 8949 adjustment applies. The sell-to-cover slice is just the first sale among many, and it happens to be the one where the basis math is automatically correct.

Holding RSUs for long-term gains is not what you think it is.

The most common reason tech employees hold their vested RSUs rather than selling them is the belief that holding qualifies the shares for long-term capital gains treatment. That belief is technically true and practically misleading. The ordinary income tax on the vest is already settled at the moment the shares hit your brokerage. From that moment forward, what you are holding is a position in your employer's stock with a cost basis equal to the vest-day fair market value, and the long-term capital gains rate applies only to the appreciation that happens after vest — exactly the same way it would apply to any other stock you bought with cash on vest day.

This is the equivalence that the "hold for long-term gains" framing obscures. Once the RSUs have vested, holding them is mathematically identical to taking an equivalent amount of after-tax cash and using it to buy your employer's stock on the open market. The capital gains treatment is a property of the holding period, not a property of the RSUs specifically. There is nothing about the historical nature of the shares as RSUs that gives them favorable tax treatment relative to a purchase you would make today.

This single test, applied honestly, settles the hold-versus-sell question for most engineers within about thirty seconds. The cash-equivalence framing strips away the loss-aversion instinct that comes from watching a brokerage balance go down when you sell, and replaces it with the question that actually matters: is your employer's stock the best use of this capital today, given everything else you own and everything else your career exposes you to? For an engineer whose salary, unvested RSUs, employer 401(k) match, and ESPP all sit at the same company, the honest answer is almost always no. The hold decision usually wins by default rather than by analysis, and the default is the bug.

What the income tax rate spread actually buys you

Engineers who hold RSUs after vest usually do so because they have heard that long-term capital gains are taxed at a lower rate than ordinary income. That is correct, but the rate difference does not work the way most people imagine. The income tax rate spread between long-term capital gains and ordinary income — roughly 17 percentage points federal at the top bracket, between the 20 percent LTCG cap under IRC Section 1(h) and the 37 percent top ordinary rate, with the 3.8 percent NIIT applying on either side — applies only to the dollars of appreciation that occur between the vest date and the eventual sale date. It does not apply to the vest-day value of the shares, because that value has already been taxed as ordinary income at the moment of vest and is no longer in the picture.

What the rate spread actually buys you, in other words, is a discount on the future appreciation of a single stock that you have chosen to keep holding for at least a year. To earn that discount, you have to accept the volatility and concentration risk of an undiversified position for the duration of the holding period. For an engineer whose career, unvested grants, and existing vested holdings already sit at the same employer, that concentration risk is meaningfully higher than the holding period of a diversified market index would carry, and the appreciation required to compensate for the risk is substantially larger than the rate spread alone suggests.

The practical upshot is that the rate spread is a real benefit but a narrow one. For households making this decision repeatedly across multiple vest cycles, the recommended discipline is to apply the cash-equivalence test at each vest and to default to selling unless there is a specific, deliberate reason to hold the position. The deeper version of the analysis, including the volatility and correlation inputs that determine the breakeven appreciation, lives in our RSU sell-versus-hold framework.

Multi-state moves and the workday allocation problem.

RSU income is sourced for state tax purposes to the state in which the work that earned the equity was performed, rather than to the state in which you live at the moment of vest. For tech employees who change residency between the grant date and the vest date — a category that grew significantly after 2020 — this distinction creates a multi-state allocation problem that the W-2 alone does not resolve. The states most likely to enforce their share of the income are the ones most likely to send an audit letter several years after the vest, by which point the documentation is harder to assemble.

California is the most aggressive jurisdiction, and the methodology it uses is the one most likely to surprise a departing resident. The Franchise Tax Board's position is codified in FTB Chief Counsel Ruling 2013-02 and the Residency and Sourcing Technical Manual, both of which apply a workday-allocation formula to the grant-to-vest period rather than looking at residency on the vest date alone. If you spent half of the period between grant and vest as a California resident and the other half as a Texas resident, California claims half the RSU income on a non-resident return for the year of vest, even though you were a Texas resident at the moment the shares actually settled into your brokerage. The California Office of Tax Appeals has affirmed this methodology in published precedent against departing residents, and a non-resident California return is not optional in the year of vest.

Skipping the non-resident return is the single most common reason that a $200,000-plus RSU vest produces a multi-year FTB audit letter three years later. By then the workday records are difficult to reconstruct, and the penalty arithmetic has had three years to compound. For multi-state engineers, the operationally correct discipline is to maintain a monthly or quarterly record of workdays per state, notify the employer's payroll team in writing on the day any residency change happens, and file the non-resident return for any year a vest hits with allocable workdays in the source state. Our California RSU guide covers the FTB workday-allocation formula in full detail, including the trailing-nexus rules that catch departing residents.

Stacking RSUs with NSOs, ISOs, and ESPP.

The decision points on a pure RSU schedule are limited, because the vesting calendar is fixed by the grant agreement and the only choice you control is what to do with the shares once they vest. The picture changes substantially when other forms of equity compensation enter the mix. Non-qualified stock options, incentive stock options, and employee stock purchase plan lots all generate taxable events that you can choose to time, and the timing decisions interact with the RSU vesting calendar in ways that compound across a multi-year career.

The first principle of coordinating these events is to avoid stacking ordinary-income recognition into the same calendar year unless you have a specific reason to do so. If your RSU vest schedule already pushes $250,000 of ordinary income into the 35 percent bracket for the year, exercising a tranche of NSOs in the same year — which generates ordinary income on the spread between strike price and current fair market value — pushes the marginal rate on the additional dollars higher than it needs to be. The same exercise in a quieter year would have been taxed at a lower bracket, and the difference compounds.

The second principle is the inverse: use low-income years for high-income decisions. The year you switch jobs and miss two vest cycles is almost always the right year to convert an old 401(k) to a Roth IRA, because the conversion itself generates ordinary income and the goal is to recognize that income at the lowest possible marginal rate. The same logic applies to a sabbatical year, a parental-leave year, or any other year in which your wage income is structurally below your normal level.

The third principle applies specifically to households holding incentive stock options alongside RSUs. An RSU vest is not itself an alternative minimum tax preference item, but the ordinary income from the vest does push your regular-tax liability higher, and a higher regular-tax liability can absorb some of the AMT credit that you generated in a prior year when you exercised ISOs. The interaction shows up only in a multi-year scenario, because neither calculation alone reveals it, and engineers holding both types of equity are the ones most likely to benefit from running an explicit projection.

For deeper treatment of the option-side mechanics, see our ISO versus NSO guide and the ESPP tax guide. Both pages cover the timing decisions and the AMT interactions in depth.

Coordinating RSU vests with mega-backdoor Roth contributions.

The mega-backdoor Roth is a feature of certain 401(k) plans rather than a Roth IRA rule, and it is one of the most powerful tax-advantaged savings tools available to high-earning employees whose plans support it. Plans that allow voluntary after-tax contributions above the IRC Section 402(g) elective deferral limit, paired with either an in-plan Roth conversion or an in-service distribution to a Roth IRA, allow a participant to push tens of thousands of dollars per year into a Roth tax wrapper that would otherwise be inaccessible at high incomes.

For tax year 2025, the Section 415(c) annual addition limit is $70,000 across all sources of contribution. Once you subtract your elective deferrals ($23,500 at the Section 402(g) limit) and the employer match, the remainder is the after-tax contribution window. At a typical FAANG-sized plan, that window runs in the $30,000 to $45,000 range per year. The mechanic depends on same-day or next-day conversion of the after-tax contribution into the Roth bucket, because any earnings that accrue on the after-tax balance before conversion become a partial taxable event when the conversion eventually happens.

RSU income does not directly affect the after-tax contribution capacity, since the Section 415(c) limit is a function of compensation rather than of which compensation has already been taxed. What RSU income does affect is the cash flow available to fund the contribution. The cleanest operational practice is to fund the after-tax bucket immediately after each RSU vest settles, since that is the moment your liquidity is highest and your bracket has already absorbed the vest income. Aligning the contribution timing with the vest schedule turns an ad-hoc savings decision into a repeatable calendar event.

One final detail to confirm with HR before designing the contribution schedule: the Section 401(a)(17) compensation cap of $350,000 for 2025 stops the employer match calculation once your year-to-date eligible compensation reaches the cap. If the plan calculates the match on a pay-period basis rather than a true-up basis, late-year contributions may miss the match entirely, which is worth knowing in advance.

When the RSU position itself becomes the problem.

The conventional threshold at which a single-stock holding becomes a concentrated position is roughly 10 to 15 percent of liquid net worth in a single name. For tech employees with vested RSUs at a long-tenured employer, the realized number frequently lands somewhere between 40 and 80 percent, because the same employer that pays the salary also issues the equity that has compounded across multiple grant cycles. The distinction between an everyday equity holding and a concentrated position matters because the planning problem is structurally different on the two sides of the threshold. Below the threshold, the right question is how to optimize the next vest in isolation. Above it, the right question is how to risk-manage the position you already have, while respecting the tax constraint that selling the position outright would generate a sizeable capital gains bill.

Quantifying the today-bill — what you would owe if you liquidated the entire position this year — is the prerequisite for every subsequent decision. The calculation uses the long-term capital gains rates under IRC Section 1(h), the 3.8 percent NIIT under IRC Section 1411, and the applicable state add-on. Once the today-bill is on paper, the discussion turns to strategies that move the position toward diversification without triggering the entire bill in a single year. Five non-sale paths cover most of the practical ground:

Each of these strategies handles a different slice of the concentration problem, and the right answer for a household with a position above $1 million is usually some combination of two or three of them layered across the same tax year. A direct-indexing program can generate the realized losses that offset the gains released through a Section 351 conversion, for example, while a charitable remainder trust handles the highest-basis tranche the household would have donated anyway. The layered structure produces a smoother glidepath toward diversification than any single strategy used alone. For a deeper treatment of when the layering math becomes worth pursuing, see our concentrated-stock guide.