You know staking rewards are taxable. The harder question is when — in which tax year, at what moment, and under which doctrinal framework. Get the timing wrong and you either overstate income in an early year (a real problem if FMV was higher then) or understate it in a later year (an audit risk). The IRS has given partial answers; some important questions remain open. This piece maps the territory as clearly as the current guidance allows.

This is an educational overview, not legal or tax advice. The timing of staking income has unresolved dimensions under current law. For positions on your specific return, consult a qualified tax professional.

Two frameworks, one ruling

Before Rev. Rul. 2023-14, taxpayers and advisors drew on two overlapping frameworks for digital-asset income timing. Understanding both is useful because they reach different results in some fact patterns — and current guidance doesn't cleanly resolve which one governs every scenario.

Dominion and control is the standard the IRS applied in Rev. Rul. 2023-14 (2023-33 I.R.B. 484, Aug. 14, 2023). The ruling establishes that PoS staking rewards are includible in gross income under IRC §61 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. This framing has roots in property-law thinking: you recognize income when you acquire genuine ownership of an asset — when you can dispose of it, transfer it, or otherwise control it as your own.

Constructive receipt is a complementary doctrine under IRC §451(a) and Treas. Reg. §1.451-2(a). It holds that income is constructively received in the taxable year during which it is credited to the taxpayer's account, set apart for the taxpayer, or otherwise made available so the taxpayer may draw upon it at any time — unless the taxpayer's control of its receipt is subject to substantial limitations or restrictions. The Supreme Court's articulation in Commissioner v. Indianapolis Power & Light Co., 493 U.S. 203 (1990), is the operative modern statement: the key question is whether the taxpayer has "complete dominion" over the funds, not merely whether the taxpayer has made use of them.

The two frameworks converge in most standard situations. For Ethereum staking, they can diverge because of the mechanics that sit between the consensus layer (where validator balances accrue) and the execution layer (where ETH actually arrives in a spendable address).

What Rev. Rul. 2023-14 actually says

The ruling addressed Situation 1: a cash-method taxpayer who validates PoS blockchain transactions and receives rewards. The IRS's holding: those rewards are includible in gross income in the taxable year in which the taxpayer gains dominion and control, at FMV as of that date and time. The ruling explicitly covers both direct staking and exchange-facilitated staking.

The ruling does not draw a bright line between the consensus layer and the execution layer. It establishes the standard — dominion and control — and leaves the application to the facts.

For Ethereum validators with 0x01 credentials (the standard setup since Capella, April 2023), the most natural "dominion and control" event is when the consensus-layer sweep credits the execution-layer withdrawal address. That is the moment the ETH becomes spendable: the staker can transfer it, use it, or do anything else an owner can do. The withdrawal credit is the clearest dominion-and-control event in the Ethereum withdrawal architecture.

The ruling is grounded in IRS Notice 2014-21, which established that virtual currency is property for federal tax purposes. Rev. Rul. 2023-14 builds on that foundation: once you accept that staking rewards are property you receive, the question becomes when you receive them in the legally operative sense.

The per-epoch vs. withdrawal-sweep divide

The doctrinal frame produces two distinct practitioner readings, each defensible under the existing authority.

Withdrawal-time (the common CPA default): Income is recognized when ETH is credited to the EL withdrawal address. The validator balance that accrues on the consensus layer is not yet income — it is locked behind the validator-exit and withdrawal-queue mechanics of the protocol, which take time and cannot be bypassed. Under this reading, the withdrawal queue constitutes a "substantial limitation or restriction" within the meaning of Treas. Reg. §1.451-2(a), so constructive receipt is deferred until the restriction is lifted and the ETH actually arrives. Recognition date is the withdrawal block timestamp; FMV is determined at that moment.

Earning-time (the per-event alternative): Income is recognized at each consensus-layer reward event — attestation, sync-committee, proposal, priority fee, MEV. The reasoning: after Capella enabled withdrawals in April 2023, the staker holds the withdrawal keys and can initiate a withdrawal at any time, subject only to the protocol queue. The queue is a procedural delay, not an absolute barrier. Under this reading, each per-event validator-balance credit is "made available without substantial restriction," and constructive receipt attaches at each event timestamp. Withdrawal events are excluded to avoid double-counting income already recognized at the per-event point.

Both readings are consistent with Rev. Rul. 2023-14 and the constructive-receipt doctrine. The difference is a factual characterization: is the Ethereum withdrawal queue mechanism a "substantial limitation or restriction" that defers constructive receipt, or is it a routine procedural delay that does not?

That factual characterization question does not have a published IRS answer. The withdrawal-time reading is conservative — it defers income to the point of spendability — and is the more common default among CPAs advising ETH validators today. The earning-time reading produces income earlier and at potentially different FMV per event. Neither position is affirmatively wrong under current guidance; neither has been validated by an IRS ruling that specifically addresses the validator-balance mechanic.

The practical implication: For most validators with 0x01 credentials operating post-Capella, the two policies converge on the same total dollar amount — the question is which tax year individual events land in and what FMV applies. For validators with long-held pre-Capella accrual (rewards that accumulated in the beacon chain before the April 2023 Capella upgrade), the divergence is significant: the withdrawal-time policy would put the entire pre-Capella accrual into whichever year the sweep was received, at FMV on the sweep date, while the earning-time policy would distribute income across the accrual period at the FMV of each epoch.

What Jarrett did — and didn't — do

Jarrett v. United States (S.D. Tenn. 2022) is the most frequently misunderstood episode in staking-tax history. Understanding what it actually decided — and what it didn't — is important for evaluating any "created property" argument.

Here is what happened. Joshua and Jessica Jarrett, Tennessee attorneys, reported 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" — analogous to a farmer's crop, which is not income until it is sold. The IRS issued a refund. The Jarretts kept the money but continued litigating to get a declaratory judgment that staking rewards are not income at receipt.

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 court ruling, a common IRS litigation tactic. No court ruled on the merits. The refund was not precedent when issued.

Rev. Rul. 2023-14 then answered the question directly: the IRS rejected the "newly created property" theory and applied the dominion-and-control standard instead. Staking rewards are ordinary income under IRC §61. No "created property" exception exists under current IRS interpretation.

As of this writing, further Jarrett litigation is ongoing but no published opinion has ruled on the merits of that case post-Rev. Rul. 2023-14. 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 guidance; the Jarrett refund is not. The "created property" theory is foreclosed under current authority — not necessarily forever, if a court rules differently, but for returns filed today.

And in June 2026, a court did rule — against the theory. In Paschall v. Commissioner, T.C. Memo. 2026-46 (June 5, 2026), the Tax Court issued its first merits decision on staking-reward taxation and held the rewards taxable on receipt under IRC §61, rejecting the same "newly created property" argument: the staker does not create the tokens, so whether they are newly created is immaterial. The court found dominion and control attached when the rewards were credited to the taxpayer's account. Paschall is a T.C. Memo. (persuasive, not binding) decided on custodial-exchange facts — the taxpayer staked Cardano through eToro, where the credited rewards were immediately tradeable with no protocol withdrawal queue in the way. That fact pattern matters for the timing discussion below: Paschall confirms that staking rewards are income on receipt and forecloses the deferral-until-sale theory, but its dominion-and-control moment (account credit, no queue) does not resolve the per-epoch-vs-withdrawal-sweep divide that solo Ethereum validators face.

The EIP-7002 open question

The Pectra upgrade (May 2025) introduced 0x02 validator credentials and EIP-7002, which allows validators to trigger withdrawals via an on-chain EL transaction rather than waiting for the automatic consensus-layer sweep. This creates a recognition-timing question that Rev. Rul. 2023-14 does not reach, because the ruling predates Pectra.

For a 0x02-credential validator, the withdrawal sequence has two meaningful timestamps:

1. The trigger transaction — the staker signs and broadcasts an EL transaction requesting the withdrawal. This is the staker's affirmative act of initiating access to the ETH.
2. The withdrawal credit — some time later, the ETH is actually credited to the EL address and becomes spendable.

The question: does dominion-and-control attach at the trigger, or at the credit?

Arguments for the trigger: the trigger is the staker's deliberate act of exercising control over the validator's accumulated balance; from that point, the withdrawal is inevitable barring an unusual protocol-level failure; the staker has made a committed decision to take possession of the ETH.

Arguments for the credit: the ETH is not yet spendable at the trigger; the protocol still governs the timing of the actual credit; the staker cannot transact with, transfer, or use the ETH until it arrives in the EL address. This mirrors the withdrawal-time argument from Situation 1.

This is an open question. Rev. Rul. 2023-14 does not address it. No subsequent IRS guidance addresses it. Both readings are plausible extensions of the existing framework; neither has been validated. For validators using 0x02 credentials, this question should be part of a counsel engagement before filing.

Why the timing question is an architecture question

The recognition-timing uncertainty has a practical implication for how staking-tax software should be designed: if the rule might change — through follow-on IRS guidance, Pectra-specific clarification, or court decisions — the software shouldn't have the policy baked in.

TrueStake's JurisdictionAdapter is built on the principle that staking receipts are immutable on-chain facts and tax recognition is computed at report time by the adapter. The raw receipt — timestamp, wei amount, event type, beacon API payload — is stored without any tax-treatment assumption. When a USJurisdictionAdapter runs, it applies the selected recognition policy (withdrawal-time as the Phase 0 default; earning-time computed in parallel as the alternative) to those stored facts and produces taxable income records.

If the rule shifts — EIP-7002 guidance, follow-on Rev. Rul. 2023-14 clarification, or legislation — the adapter's recognition policy is updated and the computation re-runs from the same stored receipts. The underlying data doesn't change; the interpretation does. No re-ingestion, no data migration, no "we'll need to redo everything."

Both the withdrawal-time and earning-time policies are computed against the same event stream so the difference in tax liability under each scenario is visible and comparable. This is the architectural decision that lets the recognition question remain open without creating a data-quality crisis.

Choosing a recognition policy in practice

For most US ETH validators filing today, the practical guidance from most CPAs is:

Withdrawal-time is the conservative default. It recognizes income at the clearest dominion-and-control moment — when the ETH arrives in a spendable address — and treats the withdrawal queue as a substantial restriction that defers constructive receipt. This is consistent with Rev. Rul. 2023-14, consistent with Treas. Reg. §1.451-2(a)'s substantial-limitations carve-out, and aligns with how most tax professionals advise today.

Earning-time is a defensible alternative. It recognizes income earlier and potentially spreads it across more tax years at per-event FMV. Some taxpayers may prefer this approach for specific reasons (including situations where earlier recognition at lower FMV produces a better outcome). It is not affirmatively inconsistent with Rev. Rul. 2023-14; it is the more aggressive of two plausible readings.

The Jarrett position is not defensible under current guidance. Rev. Rul. 2023-14 directly rejects the "created property" theory. Do not rely on the Jarrett refund as authority.

EIP-7002 / 0x02 validators should seek counsel. The trigger-vs-credit question is not resolved, and the stakes are real if FMV at the trigger and FMV at the credit differ.

Keep records at both the recognition moment and the withdrawal event. Even if you use withdrawal-time, keeping the per-event beacon API data means you can recompute under the earning-time policy retroactively if guidance changes.

What defensible documentation looks like

Whatever recognition policy you choose, the requirement is the same: a chain of custody from on-chain fact to tax return line item, with citations at every step.

For each recognized income event that means: the validator index and epoch (or block, for proposals and MEV), the wei-denominated reward amount from the beacon API response, the timestamp of the recognition event, the ETH-USD fair-market value at that timestamp (methodology documented and applied consistently, per Notice 2014-21), and the resulting dollar amount of ordinary income.

For withdrawal-time filers, the additional link is the execution-layer transaction that credits the EL withdrawal address, and a reconciliation confirming that the beacon-derived amount and the on-chain receipt match within tolerance. The reconciliation is what converts a computation into a verifiable fact.

The timing question — which year, which event — has doctrinal complexity. The documentation requirement is simpler: show your work, cite the primary sources, and store the raw data so the computation can be reproduced. That is what an examiner actually asks for. The primary sources are Rev. Rul. 2023-14 (for the dominion-and-control standard and FMV methodology), IRC §451(a) and Treas. Reg. §1.451-2(a) (for the constructive-receipt framework), and Notice 2014-21 (for virtual currency as property). Document which policy you applied and why, and keep that documentation with the return.

Policy coverage reflects the research library as of September 2026. Primary sources: IRS Rev. Rul. 2023-14 (Aug. 2023), Paschall v. Commissioner, T.C. Memo. 2026-46 (June 5, 2026), IRC §451(a), Treas. Reg. §1.451-2(a), Commissioner v. Indianapolis Power & Light Co. (U.S. 1990), IRS Notice 2014-21 (Mar. 2014), Jarrett v. United States (S.D. Tenn. 2022). EIP-7002 recognition timing is an open question under current guidance — no IRS ruling or court decision addresses the Pectra fact pattern. This article reflects educational research, not legal or tax advice. Consult a qualified tax professional for positions on your specific return.