The short answer
Bay Area CPA guide to crypto staking taxes: when ETH, SOL, Lido rewards are taxable under Rev. Rul. 2023-14, dominion-and-control, FBAR for San Jose validators.
A Bay Area validator who earns 25 ETH of staking rewards in a year at an average $3,800 price triggers roughly $40,000 to $48,000 of federal-plus-California ordinary tax in the year of receipt, even if not a single token has been sold. If you have been staking ETH through Lido, running your own validator on Solana from a San Jose home setup, or earning rETH from Rocket Pool, the IRS has a clear answer for you in 2026: those rewards are ordinary income at the moment you can use them, measured at fair market value in USD on the day you receive them. The fact that you have not sold a single token does not delay the tax. The fact that the price has crashed since the reward hit your wallet does not reduce it.
That is the rule under Rev. Rul. 2023-14, the 2023 IRS pronouncement that closed a multi-year argument over whether staking rewards are "newly created property" (like a baker baking bread) or gross income at receipt. The IRS picked income. The Jarrett case in the Middle District of Tennessee tried to push back. The taxpayers won a refund but lost the precedent fight, and the ruling has stood ever since.
This guide walks through what staking rewards are, when the taxable event occurs, the special wrinkles around liquid staking tokens (LSTs) and restaking, what to do if you run your own validator with a multi-week unbonding period, and how to actually file the income on your 1040. If you stake meaningful amounts of ETH, SOL, ATOM, DOT, or anything that issues protocol rewards, our crypto tax team sees these returns weekly.
Proof-of-stake networks pay validators and delegators new tokens for posting attestations, proposing blocks, and securing the chain. The economic mechanism varies by protocol (Ethereum issues new ETH plus a slice of priority fees and MEV; Solana pays in SOL inflation; Cosmos pays in ATOM at a variable rate), but the federal tax characterization is the same: when the rewards become yours, you have income.
That income falls under IRC Section 61, the catch-all "gross income from whatever source derived." It is taxed at your ordinary marginal rate, the same brackets as W-2 wages. For a Bay Area software engineer in the 32% federal bracket and the 9.3% California bracket, that is north of 41 cents on the dollar before any payroll-style add-ons.
The Internal Revenue Code does not have a dedicated staking provision. The framework borrows from old "treasure trove" and "found money" cases (Cesarini v. United States, the piano-found-cash case from 1969) and from prior crypto guidance on hard forks and airdrops (Rev. Rul. 2019-24). Rev. Rul. 2023-14 finally extended that logic to staking by name.
The IRS held in July 2023 that a cash-method taxpayer who stakes native tokens of a proof-of-stake protocol and receives validation rewards must include the fair market value of those rewards in gross income in the taxable year in which the taxpayer gains dominion and control over them.
Dominion and control is the operative concept. The IRS defines it as the moment the taxpayer can sell, exchange, or otherwise dispose of the rewards. For an Ethereum solo staker after the Shapella upgrade, that means the moment a partial withdrawal hits the withdrawal credential address. For a Coinbase or Kraken customer staking through the exchange, that means the moment the rewards post to the trading balance and are spendable.
The taxable amount is the USD fair market value at that exact moment. If you receive 0.05 ETH on a day when ETH is trading at $3,800, you have $190 of ordinary income, full stop. That $190 becomes your cost basis in the new 0.05 ETH. If you later sell at $4,500, you have a $35 capital gain (subject to short- or long-term treatment based on the holding period from the income date). If you sell at $3,000, you have a $40 capital loss.
This is the doctrine: income at receipt at FMV, then capital character on the subsequent disposition. The two events are distinct.
Joshua and Jessica Jarrett, Tezos stakers in Nashville, paid roughly $9,400 of income tax on 8,876 XTZ they earned by baking in 2019. They then filed a refund claim arguing the rewards were newly created property, analogous to a farmer growing crops or a baker baking bread, and that income should not arise until the property is sold. The taxpayer creates the asset; the asset is not received from anyone.
The DOJ Tax Division agreed to refund the tax in 2022 rather than litigate. The Jarretts refused the refund, wanting a precedential ruling on the merits. The court dismissed the case as moot once the refund was tendered. No binding precedent was created. The taxpayers got their money back. The IRS got its position untouched.
The IRS then issued Rev. Rul. 2023-14 the following year to lock in the dominion-and-control framework. The Jarretts filed a second refund suit for tax years 2020 and 2021. That second case (Jarrett II) is winding through the same Middle District of Tennessee court. As of mid-2026, no decision has reversed the IRS position. Practitioners treat Rev. Rul. 2023-14 as the operative authority.
Translation for Bay Area stakers: you can take the Jarrett position on a return, but you are taking a non-frivolous-but-against-the-government stance that requires Form 8275 disclosure and a willingness to litigate. Most clients we advise file under the IRS framework and preserve cash for the more productive tax fights.
The hardest practical question in 2026 is when dominion and control arises for liquid staking tokens. When you deposit ETH into Lido, you receive stETH (a rebasing token whose balance grows daily) or wstETH (a wrapped non-rebasing version). When you deposit into Rocket Pool, you receive rETH (a non-rebasing token that appreciates against ETH over time). When you restake through EigenLayer, you may receive an LRT like eETH or rsETH layered on top of stETH.
Three plausible timing positions exist:
The IRS has not issued specific guidance on LSTs. Most practitioners, ourselves included, default to position 1: treat the rebase or appreciation as ordinary income at the time the LST holder's balance or unit value increases. The LST is freely tradeable on Curve and Uniswap; dominion and control exists from day one. Crypto-tax software (Koinly, CoinTracker, TaxBit, ZenLedger) follows this convention by default, which means most retail stakers are reporting under position 1 whether they realize it or not.
Position 2 has logical appeal but no IRS support and exposes you to a substantial-understatement penalty if challenged. We do not recommend it without a documented Form 8275 position and a real willingness to defend it.
EigenLayer and similar restaking protocols let stakers re-pledge their LSTs to secure additional services in exchange for additional rewards (points, native tokens, sometimes both). Each new reward stream is its own income event at FMV when received. The points are the gray area: most practitioners treat points as zero-basis property with no income event until they are converted to a token with a market price. Once converted, full FMV at conversion is ordinary income.
If you run your own Ethereum validator (32 ETH staked, your own infrastructure), Solana validator, or Cosmos validator, the analysis is structurally the same but the timing facts differ.
Post-Shapella Ethereum validators receive partial withdrawals to the withdrawal credential address roughly every few days. Each partial withdrawal is income at the FMV of the ETH on the date it sweeps. Block proposal rewards (including MEV) are income on the day the block is proposed and the reward credited to the fee recipient address.
For protocols with multi-week unbonding (Cosmos at 21 days, Polkadot at 28 days), a more aggressive position is that the rewards are not subject to dominion and control until the unbonding period completes. The argument: during unbonding the tokens are slashable and cannot be transferred. The IRS has not blessed this position. Conservative practice is to recognize income at the protocol-level attestation or epoch reward event, not at the end of unbonding. If you want to take the unbonding-delay position, document it with Form 8275 and a memo.
Slashing losses are deductible as IRC §165 losses, but only against the basis you have in the slashed tokens. If your validator was slashed for 0.5 ETH that you had already recognized as income at $3,800, you have a $1,900 loss to offset gains. After 2017's TCJA, miscellaneous itemized deductions for individual investors were suspended through 2025; the One Big Beautiful Bill Act (OBBBA, Public Law 119-21, signed July 4, 2025) made that suspension permanent. Either way, the slashing loss generally only helps if it is investment-related and qualifies as a capital loss on disposition rather than a casualty loss.
Running through a year of staking rewards for your own return? We reconcile Koinly and CoinTracker exports against on-chain data for Bay Area validators and DeFi-heavy traders every week. Schedule a complimentary consultation.
The most common mistake we see on crypto-staking intakes: a Bay Area engineer self-files three years of staking activity through a single Koinly export without checking that the dominion-and-control timing is right. The result is usually a $30K to $80K understatement of ordinary income spread across the years, which surfaces in 2025 and later when 1099-DA broker reports start matching against returns. The IRS now receives broker-side data on every exchange-staked reward; mismatches are flagged automatically.
The second pattern: DIY filers who skip FBAR on a Binance international account because "the wallet is mine, not the exchange's." A custodial exchange account holding any staking position above $10K aggregate triggers FBAR, and the non-willful penalty is $16,117 per missed year (2025 inflation-adjusted figure). DIY filing software does not flag the FBAR obligation. The cost of a missed three-year FBAR window can clear $50K in penalties alone, before any tax adjustment.
For an individual taxpayer who is not in the trade or business of staking, the mechanics are:
Practical software workflow: connect your wallets (or exchange API for custodial stakers) to Koinly, CoinTracker, or TaxBit. The tool pulls every reward event with the timestamp and the spot price. Export the year-end report. The "Income Report" line is your Schedule 1 number. The "Capital Gains Report" is your Form 8949 input.
If you use Coinbase or Kraken for staking, expect a Form 1099-MISC for the income portion. The exchange's FMV may not match your software's; reconcile and use the higher of the two (or the one with the better documentation trail) for your return. The IRS now receives digital asset broker reporting on Form 1099-DA starting with 2025 transactions, so the match-game between exchange-reported and taxpayer-reported numbers has gotten much tighter.
If staking rises to the level of a trade or business under IRC Section 162, the income is subject to self-employment tax (15.3% on the first ~$176,100 of net SE earnings using the 2025 SS wage base; the 2026 figure is subject to the fall 2025 Rev. Proc. announcement), 2.9% above that, plus 0.9% Additional Medicare on high earners. It is also potentially eligible for QBI deduction under §199A and for retirement-plan contributions like a solo 401(k).
The IRS has not issued guidance on when staking crosses the line. Practitioner consensus borrows from the existing trade-or-business analysis (Comm'r v. Groetzinger, 480 U.S. 23) and looks at:
A retail user who delegates to Lido and earns $3,000 of stETH rebase income in 2026 is not in a trade or business. A user running 30 ETH validators with co-located hardware, paying for MEV-Boost relays, and offering delegation services to friends and family probably is. Most of our clients sit in between and report on Schedule 1 (no SE tax). The line is fact-specific. We typically run a written memo for clients above $50,000 of annual staking income or anyone running their own validator infrastructure.
If you stake through an offshore exchange (Binance.com international, Bybit, OKX non-US, KuCoin), those accounts are foreign financial accounts and trigger FBAR (FinCEN Form 114) and potentially Form 8938 reporting once thresholds are met. The FBAR threshold is the aggregate of $10,000 across all foreign accounts at any point in the year. Form 8938 thresholds start at $50,000 for single filers living in the US.
For a deeper walk-through of those rules, see our 2026 FBAR filing deadline guide.
Self-custody staking (running your own validator, holding ETH in MetaMask, staking through Lido from your own wallet) is generally not an FBAR-reportable activity. FinCEN's proposed rule for unhosted wallets has been in limbo since 2020 and has not become final. The current FBAR instructions speak in terms of "financial accounts" with a custodian, which a self-custody wallet is not.
This split matters. A Bay Area engineer who stakes 64 ETH on Lido from a MetaMask wallet has zero FBAR exposure. The same engineer who moves the position to Binance international for higher yield has full FBAR exposure on the account balance, plus an FBAR filing obligation, plus a real audit risk.
California treats staking income the same way the IRS does: ordinary income at FMV at receipt, ordinary state rate up to 13.3% (12.3% top bracket plus 1% Mental Health Services Act surcharge that applies on taxable income above $1M single / $2M MFJ; the combined top rate reaches 14.4% under certain higher-bracket scenarios). No special crypto regime. California does not recognize §1031 like-kind exchange treatment on crypto and never has (it was eliminated for personal property at the federal level by TCJA anyway).
If you moved to or from California during the year, sourcing matters. California sources the income to the state where you were a resident at the time of receipt. A move from California to Texas mid-year does not retroactively make pre-move staking income Texas-sourced.
The IRS standard is dominion and control, not physical claim. If the rewards are credited to your account or wallet address and you have the ability to claim them at will (the typical case for ETH partial withdrawals post-Shapella and for most exchange staking), you have income. If a protocol has a true lockup that prevents claim (some early staking programs, certain validator-bonded positions), an unclaimed-rewards argument is stronger. Most clients have income regardless of whether they have clicked the "claim" button.
Use the best available pricing source: a thin DEX market, a CoinGecko aggregated mid-price, or the issue price from the protocol. Document the source and apply it consistently. If no market exists at all, FMV is genuinely indeterminate and an open-tax-year position can be taken to defer income until a market develops. Document the position contemporaneously.
Both are ordinary income at FMV when you have dominion and control. The framework is the same. Rev. Rul. 2019-24 covers airdrops resulting from hard forks specifically (not all airdrops in general); Rev. Rul. 2023-14 covers staking. The income character, valuation timing, and basis-becomes-FMV rules all line up. The difference is only how the tokens reach you. For airdrops outside the hard-fork fact pattern, the IRS has not issued definitive guidance, but the dominion-and-control framework still controls in practice.
Yes, but the path matters. A slashing loss against tokens you have already recognized as income is a §165 loss against your basis. If the loss arises in a trade or business, it is ordinary. If you are an individual investor, it generally becomes a capital loss on disposition rather than a casualty loss (casualty losses for individuals are now severely limited under TCJA and OBBBA). Document the slashing event, the wallet address, and the loss amount.
Almost certainly not. A passive Lido delegation does not look like a trade or business under the Groetzinger factors. You are not holding out staking services to the public, you are not running your own infrastructure, and the activity is not regular and continuous beyond pressing "deposit" once. Schedule 1 reporting without SE tax is the standard treatment. Confirm with a tax professional if your staking position exceeds a few hundred thousand dollars or if you are also operating your own validators.
Same framework. Ordinary income at FMV at receipt, basis becomes FMV. Stablecoins have a near-1:1 USD value so the math is simple, but the reporting is identical. If you are running validator-as-a-service for clients who pay you in USDC, that almost certainly is a trade or business and SE tax applies.
Staking returns get complicated fast. A solo ETH validator who restakes through EigenLayer, earns points and LRTs, and occasionally hits a slashing event has a return with three layers of crypto-tax complexity stacked on top of a normal W-2 1040. The cost of doing this with an unqualified preparer (or doing it yourself with a tool you do not fully understand) is paying ordinary tax on income you should have characterized as capital, or worse, missing FBAR and getting a six-figure penalty letter from FinCEN.
At Silicon Valley Tax we work with validators, DeFi-heavy traders, and tech employees with crypto compensation across the Bay Area. We reconcile Koinly and CoinTracker exports against on-chain data, document dominion-and-control positions, plan for the California layer, and handle the FBAR and 8938 stack when offshore exchanges are involved. See our full services menu or our who we serve page for context.
Schedule a complimentary consultation and we will walk through your specific staking setup, dominion-and-control posture for any LST positions, and what your 2026 return is going to look like before April panic sets in.
Get your dominion-and-control positions documented, your basis tracked, and your FBAR exposure assessed before tax season. We work with crypto-heavy returns every week.