Staking Income Tax Character: How the IRS Treats Rewards

By the Finance Fundamentals Tax Desk. Reviewed by Marguerite Aldous, CPA, for technical accuracy. This article addresses United States federal tax rules only; state tax treatment of digital assets varies and is not covered here.

Why this question is still contested

The Jarrett refund fight

A single Tennessee couple has spent six years arguing that the IRS has this backward. Josh and Jessica Jarrett staked Tezos tokens, received 8,876 new XTZ in 2019, and paid tax on them as ordinary income. Then they asked for a refund, arguing that newly created property, like a farmer’s crop or a manufacturer’s finished good, isn’t income until it’s sold.

The Department of Justice did not want to litigate that theory. In December 2021, the government offered the Jarretts a full refund of $3,793 plus interest for the 2019 tax year.

Most taxpayers would have taken the check. The Jarretts refused it, calling the offer an attempt to dodge a precedent-setting ruling.

In September 2022, the U.S. District Court for the Middle District of Tennessee dismissed the case as moot anyway. You can’t sue over a refund the government already agreed to pay, the court reasoned, regardless of why you don’t want it.

That left the underlying legal question unanswered. The Jarretts tried again.

On October 10, 2024, they filed a second suit covering their 2020 tax year. It repeats the argument that staking rewards resemble a stock split or a newly grown crop, not a wage or a dividend. As of this writing, no court has resolved that claim on the merits.

What Rev. Rul. 2023-14 actually says

Meanwhile, the IRS’s own position has not moved. Revenue Ruling 2023-14 states plainly that a cash-method taxpayer who stakes cryptocurrency and receives validation rewards must include those rewards in gross income.

The required amount is “the fair market value of the validation rewards received,” counted “in the taxable year in which the taxpayer gains dominion and control.”

That phrase, dominion and control, comes straight out of Commissioner v. Glenshaw Glass Co., the 1955 Supreme Court case that defined income. Income, the Court held, means “undeniable accessions to wealth, clearly realized, and over which the taxpayers have complete dominion.”

The IRS is not inventing a new theory for crypto. It’s applying a seventy-year-old doctrine to a new kind of property.

My own read: the IRS position is the more defensible one under existing law. It still creates a genuine cash-flow problem for validators who receive illiquid governance tokens they can’t easily sell to cover the tax bill.

The Jarretts’ stock-split analogy is intuitively appealing but doesn’t map cleanly onto how staking rewards actually work. A stock split doesn’t hand you new voting power or new economic rights you didn’t already have a claim to, while a staking reward often does.

Reasonable tax professionals disagree on this. Until a court rules on the merits, that disagreement matters for anyone deciding how aggressively to plan around it.

Staking versus mining: two different mechanics, similar tax outcomes

Mining and staking solve the same problem, getting a decentralized network to agree on which transactions are valid, through different mechanics, and the IRS treats the income they generate almost identically.

Mining: proof-of-work rewards

Mining, used by Bitcoin and similar proof-of-work networks, requires dedicated computing hardware racing to solve a cryptographic puzzle. Whoever solves it first gets to add the next block and collects a block reward plus transaction fees.

IRS Notice 2014-21, issued March 25, 2014, addressed this directly: when a taxpayer successfully mines virtual currency, the fair market value of the coins as of the date of receipt is includible in gross income. If the mining rises to the level of a trade or business, the net earnings are also subject to self-employment tax.

Staking: proof-of-stake rewards

Staking, used by Ethereum and most newer networks, works differently. Instead of burning electricity, validators lock up (or “stake”) existing coins as collateral and are chosen, often by a pseudo-random algorithm weighted by stake size, to propose or attest to new blocks. Rev. Rul. 2023-14 addresses this scenario specifically and reaches the same bottom-line answer as the mining notice: ordinary income at fair market value, recognized when the taxpayer can actually use, sell, or transfer the reward.

The economics behind the two also differ in a way that matters for planning, even though the tax result lands in the same place. A mining operation has real, deductible operating costs: electricity, hosting, hardware depreciation, cooling.

A staking validator’s costs are mostly opportunity cost (capital that could be doing something else) and, in some setups, a commission paid to a staking service provider. That asymmetry means two taxpayers can earn the same dollar amount of ordinary income from mining versus staking and end up with very different net positions after deductions.

The practical difference between the two shows up in a few distinct places:

• Self-employment tax exposure is more commonly triggered by mining operations run as a business than by passive staking through a wallet, though a large-scale validator operation can cross that line too.
• Timing can differ because staking rewards are frequently subject to lock-up or unbonding periods that mining rewards don’t have, which pushes the dominion-and-control date later.
• Hardware deductions (depreciation, electricity, facility costs) are far more relevant to mining than to staking, where the “equipment” is often just capital already sitting in a wallet.

Timing: when does the clock actually start?

This is where most reporting mistakes happen. The trigger isn’t “when the blockchain creates the reward.” It’s when you gain dominion and control: the ability to sell, exchange, or otherwise dispose of it.

Custodial versus non-custodial staking

For a validator running their own node with no lock-up, that’s usually the moment the reward lands in a wallet they control. For staking done through an exchange or a liquid-staking protocol, it can be later, sometimes much later, if rewards accrue to a locked position and aren’t claimable until a specific unstaking window opens.

Rev. Rul. 2023-14’s own example turns on exactly this point.

The taxpayer in the ruling had staked through a service where rewards were locked. The IRS held that income wasn’t recognized until the lock-up ended and the taxpayer could actually move the tokens.

Custodial staking-as-a-service arrangements deserve special attention here. When an exchange stakes your coins on your behalf and credits your account with rewards, the dominion-and-control test still applies.

But the facts that determine the answer, whether you can withdraw today, whether there’s a waiting period, how often the exchange batches payouts, live in the platform’s terms of service, not in the blockchain’s code.

That’s a gap in how most guidance discusses this. It assumes a technically sophisticated validator reading protocol documentation. In practice, most retail stakers are just reading a webpage that says “rewards distributed weekly” and nothing else.

Get that document and keep it. If the IRS asks when you had access, you’ll need it.

Native, non-custodial staking adds its own wrinkle. Many proof-of-stake networks impose an exit queue or unbonding window before a validator can fully withdraw staked capital. That window can stretch longer depending on how many other validators are exiting at the same time. Whether that queue delays recognition of already-earned rewards, as opposed to the underlying staked principal, is exactly the kind of fact-specific question Rev. Rul. 2023-14 is meant to answer, and it’s worth documenting separately from the exit of your principal stake.

Basis: what happens when you later sell

Once a staking or mining reward has been taxed as ordinary income, its fair market value at that moment becomes your cost basis in the tokens. That basis matters for exactly one thing going forward: figuring the gain or loss when you eventually sell, swap, or spend those specific coins.

Sell the tokens for more than that basis, and the difference is a capital gain: short-term if you held the coins for a year or less after receipt, long-term if you held them longer. Sell for less, and it’s a capital loss you can use to offset other gains, subject to the usual capital-loss limitations.

This is the same two-step pattern the tax code applies to compensation paid in property generally. Ordinary income at receipt, then capital gain or loss at disposal, measured from that receipt-date value.

Holding-period tracking gets genuinely hard when rewards arrive daily or even hourly, which is common on many proof-of-stake networks.

Each reward is technically its own lot with its own basis and its own holding-period clock.

Lumping a year’s worth of staking rewards into one average-cost figure is easier. It isn’t what the law asks for, though, and it can misstate whether a later sale is short-term or long-term.

Why your 1099 forms probably won’t cover this

Form 1099-DA, the new digital-asset broker reporting form, does not solve the staking-income problem, and it isn’t designed to. Final regulations require brokers to report gross proceeds from digital-asset sales and exchanges starting with 2025 transactions, filed in early 2026, with cost-basis reporting phasing in for 2026 transactions onward. That form covers dispositions, the sale side of the ledger.

Staking and mining rewards sit on the other side of the ledger, as income, and Form 1099-DA generally does not capture them. Some payers may separately issue a Form 1099-MISC or similar form for reward income, but coverage is inconsistent across exchanges and non-existent for self-custodied validators and home miners.

The IRS also granted transition relief in Notice 2025-7. For the 2025 calendar year, taxpayers with custodial brokerage accounts could identify which specific units they sold using their own books and records, rather than relying solely on the broker’s default accounting method.

The upshot: nobody is going to hand you a clean summary of your staking income the way a brokerage sends a consolidated 1099 for stock dividends. Until broker reporting catches up, and it may never fully catch up for non-custodial staking, the recordkeeping burden sits entirely with the taxpayer.

In my assessment, this gap is a bigger practical risk than the unresolved Jarrett litigation. A court fight over legal theory takes years and affects how future rewards might be taxed. A missing basis record affects a return you have to file this April, with penalties that don’t wait for test cases to resolve.

A worked example

Say a taxpayer, call her Mara, runs a validator on a proof-of-stake network. On March 15, 2026, the protocol credits her wallet with 2 SOL in staking rewards, and she has immediate, unrestricted access to them. SOL is trading at $140 that day.

Mara reports $280 of ordinary staking income for 2026 (2 SOL at $140) and that $280 becomes her cost basis in those two coins. On November 2, 2026, she sells both coins for $210 each, or $420 total.

She has held the coins for well under a year. The $140 difference between her $420 in proceeds and her $280 basis is a short-term capital gain, taxed at her ordinary income rate and separate from the $280 she already reported back in March.

Two separate tax events, two separate characters of income.

Miss the first one and you’ve understated ordinary income for 2026. Miss the second and you’ve either overstated a loss or understated a gain, depending on which way the price moved.

Staking, mining, airdrops, and hard forks compared

These four events get lumped together constantly in casual conversation, but the IRS doesn’t treat them identically in every respect, even though all four ultimately produce ordinary income at fair market value on receipt. The table below lines up the current guidance.

Note the last row. A hard fork by itself, a protocol split with no new tokens landing in your wallet, isn’t a taxable event.

It’s only when the fork comes bundled with an airdrop of new coins, and you actually receive and control them, that income shows up. People routinely conflate “there was a hard fork” with “I owe tax,” and the two are only connected when coins actually change hands.

Where DeFi complicates the picture

Staking rewards earned directly through a DeFi protocol, rather than through a centralized exchange or a simple validator setup, raise two separate sets of questions at once. One is how the reward itself gets characterized under the dominion-and-control test. The other is how the surrounding transaction, a token swap into a liquidity pool, a wrapped-token conversion, a governance-token claim, gets taxed under the broader rules for DeFi transaction tax treatment.

A liquid-staking derivative token, for instance, might itself be treated as a separate disposition event when it’s minted, independent of whatever staking reward eventually accrues underneath it. Untangling the two requires looking at both frameworks side by side, not assuming staking guidance alone covers everything happening inside a DeFi position.

Common mistakes

A short list, not the usual three-item version people expect:

1. Waiting for a cash-out to report anything. Ordinary income is triggered by dominion and control, not by converting to dollars. Rewards sitting untouched in a wallet still generate a tax bill for the year you could have accessed them.
2. Averaging basis across a year of daily rewards. Each reward is its own lot with its own receipt-date value and its own holding-period clock; collapsing them into one number understates the recordkeeping the law actually requires.
3. Treating a hobby-scale mining rig like a business, or a business-scale one like a hobby. The self-employment tax and deduction consequences swing hard depending on which side of that line you’re on, and the line is about facts and circumstances, not the size of your electric bill alone.
4. Ignoring lock-up periods when custodial staking is involved. If a platform’s terms say rewards aren’t withdrawable for 30 days, that delay can genuinely move the recognition date, but only if you can document it.
5. Assuming a hard fork always means income. No new coins, no income, regardless of how much press coverage the fork gets.

Recordkeeping checklist

Keep this running for every wallet and every protocol you interact with:

• Date and, where possible, time of each reward’s receipt
• Fair market value in USD at that exact date/time, plus the pricing source you used
• Whether the reward was immediately accessible or subject to a lock-up, and the terms governing that lock-up
• Wallet address and protocol or exchange involved
• A running basis ledger, lot by lot, tied back to each reward’s receipt-date value
• Sale or disposal records: date, proceeds, and which specific lot was sold
• Business-activity documentation (equipment costs, electricity bills, Schedule C filings) if mining or validating rises to a trade or business

None of this is exotic. It’s the same discipline brokerage accounts have handled automatically for decades.

Crypto just hasn’t caught up yet. The burden sits with the taxpayer.

Key takeaways

Frequently asked questions

Is staking income taxed twice, once as income and once as a capital gain?
No. The reward’s value is taxed once as ordinary income at receipt. Any later gain or loss on sale is measured from that same value. You’re never taxed twice on the same dollars; you’re taxed on two different things, the reward itself and any price movement after you got it.

Do I owe tax on staking rewards I haven’t sold?
Generally yes, if you have dominion and control over them: the ability to sell, transfer, or otherwise use the tokens. Unsold does not mean untaxed under current IRS guidance.

How is mining income different from staking income for tax purposes?
Both are ordinary income at fair market value on receipt. Mining is more often treated as a trade or business subject to self-employment tax, particularly when run at scale with dedicated hardware. Passive staking is less commonly treated that way, though large validator operations can cross into business territory too.

What happened with the Jarrett case?
The Jarretts’ first lawsuit, over their 2019 taxes, was dismissed as moot in September 2022 after the government offered a full refund the couple refused to accept. They filed a second suit in October 2024 covering their 2020 tax year, raising the same “income at creation” argument. As of this writing, that case remains unresolved, and Rev. Rul. 2023-14 remains the IRS’s operative position.

Are airdrops and hard forks taxed the same way as staking?
The character and timing rules are similar: ordinary income at fair market value once you have dominion and control. But a hard fork with no accompanying airdrop of new coins isn’t a taxable event at all, per Rev. Rul. 2019-24.

Is staking income subject to self-employment tax?
It can be, if the staking activity is substantial and ongoing enough to constitute a trade or business rather than a passive investment, similar to how Notice 2014-21 treats business-level mining. Most casual, small-scale staking through a wallet or exchange does not typically rise to that level, but the analysis is facts-and-circumstances.

Does Form 1099-DA report my staking rewards for me?
No. Form 1099-DA reports gross proceeds (and, from 2026 transactions onward, cost basis) for digital-asset sales and exchanges handled by brokers. Staking and mining income falls outside that form’s scope, so you still need your own records of receipt dates and fair market values.

References

• Internal Revenue Service, Revenue Ruling 2023-14, issued July 31, 2023.
• Internal Revenue Service, Notice 2014-21, issued March 25, 2014.
• Internal Revenue Service, Revenue Ruling 2019-24 (hard forks and airdrops), issued 2019.
• Internal Revenue Service, “Frequently Asked Questions on Virtual Currency Transactions,” IRS.gov, page last reviewed June 28, 2026.
• Internal Revenue Service, Notice 2025-7 (digital-asset cost-basis identification transition relief for 2025).
• Jarrett v. United States, No. 3:21-cv-00419 (M.D. Tenn.), dismissed as moot September 30, 2022.
• Jarrett v. United States, second complaint filed October 10, 2024 (M.D. Tenn.), covering the 2020 tax year.
• Commissioner v. Glenshaw Glass Co., 348 U.S. 426 (1955).