If you've been running Ethereum validators long enough to file a return for tax year 2025, you've lived through the busiest stretch of US crypto-tax policymaking in a decade. A revenue ruling, a widely-misunderstood federal refund, a new information-reporting form, and two pending legislative proposals — all in the span of roughly three years. This piece maps where things stand heading into the 2026 filing year, in plain English, with citations.

This is a policy summary, not legal or tax advice. We explain what the rules say and where the unresolved questions are. For specific positions on your return, consult a tax professional.

Where we stand: Rev. Rul. 2023-14 is the controlling authority

The IRS issued Revenue Ruling 2023-14 in August 2023 (2023-33 I.R.B. 484). It addressed a simple question that had been left open for years: are PoS staking rewards includible in gross income, and if so, when?

The ruling's answer: yes, they are ordinary income under IRC §61, recognized in the taxable year in which the taxpayer gains "dominion and control" over the rewards. Fair market value is determined as of the date and time that dominion-and-control attaches. The ruling explicitly covers both direct staking (running your own validator) and exchange-facilitated staking.

The foundation for this ruling is older. IRS Notice 2014-21 — still in full effect — established that virtual currency is property for federal tax purposes. Rev. Rul. 2023-14 builds on Notice 2014-21 by applying property-treatment logic specifically to staking rewards, where "when did you receive the property?" is more complicated than it first appears.

For Ethereum validators using 0x01 credentials (the standard setup post-Capella), the clearest recognition event is when the consensus-layer sweep credits the execution-layer withdrawal address — the moment the ETH becomes spendable and fully under the staker's control. Some practitioners argue each per-epoch validator-balance credit is the operative event. Both readings are within Rev. Rul. 2023-14's scope; the withdrawal-time interpretation is the more common CPA default today.

The open question Rev. Rul. 2023-14 doesn't reach: For validators with 0x02 credentials (introduced in the Pectra upgrade, May 2025), withdrawals are not automatic sweeps — they require an explicit on-chain trigger transaction via EIP-7002. Does dominion-and-control attach when the staker signs the trigger transaction, or when the ETH actually credits the execution layer? Rev. Rul. 2023-14 predates Pectra and doesn't address this fact pattern. The answer is open.

What Jarrett did — and didn't — do

The most frequently misunderstood episode in staking-tax history is Jarrett v. United States (S.D. Tenn. 2022).

Here is what actually happened. Joshua and Jessica Jarrett, Tennessee attorneys, reported their Tezos staking rewards as ordinary income and paid tax on them. They then filed for a refund, arguing that staking rewards are newly created property and therefore not income at receipt — similar to how a farmer's crop isn't income until it is sold. The IRS issued a refund. The Jarretts kept the money but continued litigating to obtain a declaratory judgment.

What did not happen: the IRS did not concede the argument. The refund was issued without explanation — most likely to moot the case before a potentially unfavorable ruling, a common IRS litigation tactic. No guidance acknowledged that staking rewards are not income at receipt.

Rev. Rul. 2023-14 then made the IRS position explicit: it directly addressed and rejected the "newly created property" theory. The dominion-and-control standard applies; staking rewards are ordinary income; no "created property" exception exists under current IRS interpretation. The Jarrett refund was not precedent when issued. After Rev. Rul. 2023-14, the Jarrett argument is affirmatively foreclosed under current IRS guidance.

Further Jarrett litigation is ongoing, but there is no published federal court opinion ruling on the merits of that case. Coin Center and the DeFi Education Fund have published policy briefs on the competing legal theories (see citations). Filing a return that relies on the Jarrett argument as authority is not a defensible position under current IRS guidance; Rev. Rul. 2023-14 is controlling.

Paschall: the first Tax Court merits ruling

In June 2026, the question Jarrett never got a merits answer to finally reached one. In Paschall v. Commissioner, T.C. Memo. 2026-46 (June 5, 2026), the US Tax Court issued its first merits decision on the taxation of staking rewards — and held that they are includible in gross income under IRC §61 in the year of receipt, when the taxpayer gains dominion and control over them.

The taxpayer had staked Cardano through the eToro platform and received reward tokens credited to his account monthly. He argued the rewards were "newly created property" not taxable until sale — the same theory Jarrett advanced. The court rejected it directly: a staker does not create the tokens; the protocol does, and the staker merely receives them. Whether the tokens are "new" is immaterial. The court grounded its holding in §61 and the Supreme Court's Glenshaw Glass "accession to wealth" standard, found the taxpayer had dominion and control the moment the rewards were credited, and held that the platform's trading restrictions did not defeat that control.

Two limits are worth stating plainly, because most commentary you'll see this year glosses them. First, Paschall is a T.C. Memo. decision — persuasive, but not binding precedent, and the court noted that the taxpayer (proceeding without counsel) presented no expert testimony on how staking actually works. Second, the facts were custodial-exchange staking: the rewards landed in a tradeable account with nothing like Ethereum's validator-exit and withdrawal-queue mechanics between accrual and access. So Paschall squarely confirms the headline rule — staking rewards are income on receipt, and the "created property" deferral theory is a loser — but it does not resolve the finer Ethereum-specific timing question of which moment counts for a solo validator (covered in our recognition-timing article). The direction of travel, though, is unmistakable: every authority to address the question — the IRS in Rev. Rul. 2023-14 and now the Tax Court in Paschall — has landed in the same place.

1099-DA and the 2026 broker era

Form 1099-DA is the IRS's new information-reporting form for digital-asset transactions. It matters for staking income in ways that depend significantly on how you stake.

The IRS issued final regulations requiring "digital-asset brokers" to report customer transactions on Form 1099-DA beginning with transactions in tax year 2025 (reporting due in early 2026). "Broker" includes centralized exchanges, custodial wallet providers, and similar intermediaries who effectuate sales of digital assets on behalf of customers. Per the 1099-DA instructions, the regulations explicitly carve out validators and miners operating on a proof-of-work or proof-of-stake network — running a validator does not, by itself, make you a broker required to issue 1099-DAs.

What this means in practice:

For solo validators: You are not required to issue Form 1099-DA to yourself or anyone else. Your obligation is to accurately report your own staking income on your return, backed by your own records. The 1099-DA rules don't change that obligation — they just don't add an additional reporting burden on top of it.

For exchange-resident stakers: If you stake through Coinbase, Kraken, or another custodial platform that falls within the broker definition, that platform is required to issue you a 1099-DA starting for tax year 2025 transactions. The 1099-DA will cover your staking rewards paid out by the exchange. This is meaningful: it puts IRS data-matching in place. If your return doesn't match the 1099-DA the broker files for you, expect scrutiny.

The data-quality caveat: 1099-DA is new and untested at scale. Even if you receive one, the obligation to report accurately from your own records remains yours. A 1099-DA that understates your income doesn't reduce your tax; document any discrepancy.

The broader effect of 1099-DA is infrastructure: the IRS will have more data-matching capability for digital-asset income than it has ever had. The "estimate later" approach — keeping rough notes and cleaning it up when an audit arrives — gets riskier every year this infrastructure matures.

State-level developments

State income taxation of staking rewards follows federal characterization in most states: if the income is federally includible under IRC §61 (as Rev. Rul. 2023-14 establishes), it flows into the state's starting-point AGI calculation and is taxed at the applicable state rates.

A few patterns worth noting:

No-income-tax states: Texas and Florida impose no broad individual income tax. Wyoming similarly has no state income tax. For residents of those states, staking rewards produce no state income tax — though federal ordinary-income treatment under Rev. Rul. 2023-14 still applies. (Texas has a franchise tax on entities; it does not affect individual filers reporting staking income on Schedule 1 or Schedule C.)

Higher-enforcement states: California and New York conform to federal AGI without any staking-specific exclusion or modification (per CA Rev. & Tax. Code §17072 and NY Tax Law §612(a), respectively). Both states have historically enforced crypto-income compliance more aggressively than average. New York's Department of Taxation and Finance has issued advisory opinions confirming virtual currency is taxed consistently with federal treatment.

Caveat: State-specific guidance on staking rewards is sparse in most states. The conformity statutes work automatically — staking income flows through to state AGI — but most states have not addressed recognition timing, FMV methodology, or 0x02 edge cases. Consult counsel for your state's current administrative posture.

What this means operationally

The convergence of Rev. Rul. 2023-14, 1099-DA phase-in, and improving IRS data-matching capability points in one direction: record-keeping quality is no longer optional for staking income.

The practical requirements for a defensible return:

Track income by validator, not by wallet. Per-epoch or per-sweep records with FMV at each event — not an annual wallet tally. Wallet scanning misses pre-Capella accrual and misclassifies MEV tips paid to a different fee-recipient address.

Reconcile to on-chain facts. Every income figure should trace to a specific on-chain receipt or beacon API response, with a reconciliation check confirming they match within tolerance. That chain of custody is what makes a number defensible.

Cite your FMV source. Rev. Rul. 2023-14 and Notice 2014-21 require FMV at the time of receipt — a timestamped price, not an end-of-day approximation. Document the methodology and apply it consistently.

Account for 1099-DAs you receive. If you use a custodial staking platform, you will receive a 1099-DA for tax year 2025 transactions. The IRS has a matching record. Reconcile your own records against the form before filing and document discrepancies.

The policy can move; the data layer doesn't have to

Two pieces of pending legislation — S.2207, introduced by Sen. Lummis, and the Digital Asset PARITY Act discussion draft from Reps. Miller and Horsford — would substantially change staking-income recognition. S.2207 would defer recognition until sale (ordinary income at disposition; basis = $0 at receipt). The PARITY Act would provide an elective five-year deferral. Neither has advanced to committee markup as of this writing; both remain in early-introduction phase. We track both quarterly in our research library and will update here if either advances.

If either bill were enacted, it would supersede Rev. Rul. 2023-14 for the periods covered. That's a significant policy shift — and it illustrates why the data architecture matters as much as the policy.

TrueStake's JurisdictionAdapter is built on a principle from the start: store receipts as immutable facts; compute tax recognition at report time. Every staking reward is stored as a raw on-chain fact — timestamp, wei amount, event type, raw beacon API payload — with no tax-treatment assumptions baked in. When a USJurisdictionAdapter runs, it reads those facts and applies the current recognition policy (today: Rev. Rul. 2023-14's dominion-and-control standard) to produce taxable income records.

If the rule shifts — Jarrett follow-on litigation, S.2207, the PARITY Act, or anything else — we update the adapter's recognition policy and re-run the computation from the same stored receipts. The data doesn't change; the interpretation does. No re-ingestion, no data migration, no "we'll need to redo the whole thing."

That architectural decision was made deliberately, in 2025, with exactly this scenario in mind: the policy is uncertain, the facts are not. Build the system that's correct about the facts and re-runnable about the interpretation — because the interpretation is going to keep moving.

Policy coverage in this article reflects the research library as of August 2026. Primary sources: IRS Rev. Rul. 2023-14 (Aug. 2023), Paschall v. Commissioner, T.C. Memo. 2026-46 (June 5, 2026), IRS Notice 2014-21 (Mar. 2014), Form 1099-DA instructions (IRS, 2025), Coin Center policy briefs (coincenter.org/category/tax), DeFi Education Fund policy work. Pending-legislation status as of last library review. This is not legal or tax advice — consult a qualified tax professional for your specific situation.