Ethereum destroys part of the ETH paid to use its network, but that does not mean the supply of Ether must fall. New ETH is also issued to reward validators. The balance between creation and destruction determines the change in supply over a period; neither a burn counter nor a staking yield tells the whole story.
This is the distinction behind many claims that ETH is inflationary or deflationary. The fee mechanism can be operating exactly as designed while supply grows, because issuance exceeds destruction. Conversely, enough fee burning can outweigh new issuance. These are measurable outcomes over a specified interval, not permanent labels attached to the asset.
Burning changes the supply, not just the owner
A normal payment moves an asset between accounts. Burning removes it from the amount recognized by the protocol. Ethereum's base-fee mechanism does this as part of transaction processing, rather than collecting the fee in a treasury that can later spend it.
EIP-1559 introduced the execution base fee and specified that it is destroyed. The proposal's original discussion refers to miners because it predates Ethereum's transition to proof of stake. Today's validators operate under the newer consensus model, while the distinction between a burned base fee and a proposer payment remains.
The mechanism is not a company buying back its shares. There is no discretionary management team choosing the quantity to repurchase from a market. Users pay for network resources under protocol rules, and the relevant portion of those payments is removed. Describing every burn as a buyback can obscure both the source of the payment and the absence of a corporate balance sheet.
A transaction fee has more than one destination
For ordinary execution, gas measures the resources consumed by a transaction. The base fee prices each unit under the network's adjustment mechanism. A priority fee, or tip, compensates the block proposer. Only the base-fee component is destroyed through this mechanism.
The amount a wallet reserves as a maximum is not necessarily the amount finally paid. Actual gas use and the effective price matter. Our Ethereum gas guide explains fee caps, priority payments and the cost of an execution that reverts. Those user-cost questions are distinct from measuring total supply.
Consider a hypothetical transaction using 50,000 gas at a base fee of 8 gwei and a tip of 1 gwei. One gwei is one-billionth of an ETH. The base component is 400,000 gwei, or 0.0004 ETH, which is burned. The tip is 50,000 gwei, or 0.00005 ETH, transferred to the proposer. The combined execution fee is 0.00045 ETH.
If the tip rises to 3 gwei while gas use and the base fee stay unchanged, the total becomes 0.00055 ETH. The burn is still 0.0004 ETH. A higher total fee therefore does not always imply a higher burn. These are invented inputs to demonstrate the arithmetic, not a current gas quotation.
Issuance pays for consensus participation
Ethereum's issuance history explains the September 2022 transition from proof of work to proof of stake. Mining issuance ended at the Merge; validator rewards continued through the consensus layer. Historical daily-issuance examples on documentation pages should not be reused as today's measured totals.
Validators earn rewards for performing assigned consensus duties. The reward calculation depends on effective stake and overall participation, rather than promising a fixed percentage independent of network conditions. Ethereum's rewards documentation describes how larger aggregate active stake can increase total issuance while reducing the reward rate per unit staked, with other conditions held equal.
That relationship matters when interpreting staking growth. More ETH participating does not mean issuance grows in direct proportion to the deposit total. Nor can a particular customer's displayed yield be treated as the network's supply-growth rate. Provider fees, validator performance and other income components can change the customer's result.
Validator revenue is not all newly created ETH
A validator can receive newly issued consensus rewards and payments drawn from assets that already exist. Priority fees are transfers. They raise the recipient's balance without increasing aggregate supply. Keeping the categories separate prevents a validator-income figure from being mistaken for issuance.
Ethereum's MEV documentation also describes value extracted through transaction inclusion and ordering beyond ordinary rewards and gas fees. Such income is a different concept from protocol creation of ETH. A profitable ordering opportunity does not itself mean the network minted that profit.
Imagine an operator receives 2 ETH of new consensus rewards and 1 ETH from other participants during an accounting period. Its gross receipts are 3 ETH, but only 2 ETH in that example is newly created. If the network burns 2.5 ETH over the same interval, subtracting burn from the operator's combined receipts would give the wrong supply conclusion.
Our staking explainer covers participation, service arrangements and capital risk. The supply question instead asks which changes create or destroy ETH across the network, regardless of which operator receives a payment.
Compare creation and destruction over the same period
Ethereum's supply overview identifies issuance and fee burning as the two main opposing forces. A simple accounting starting point is opening supply plus new issuance minus ETH destroyed. A precise reconciliation must also account for consensus balance reductions and the methodology used by the data provider.
Suppose an imaginary network begins a day with 120 million ETH, issues 2,000 ETH and destroys 1,500 ETH, with no other changes in the example. It ends with 120,000,500 ETH: supply increased by 500 ETH even though the burn mechanism removed a meaningful quantity.
If destruction instead reaches 2,500 ETH on the same issuance assumption, ending supply is 119,999,500 ETH. The supply falls by 500 ETH. The calculation says nothing about the market price in either scenario. Buyers and sellers can value an asset differently even when its quantity is known.
Period selection matters too. A day of net destruction can occur inside a month of net issuance. Comparing a daily burn number with an annual issuance estimate mixes denominators. Any inflation or deflation rate needs an interval, a starting-supply reference and a clear definition of included changes.
Rollup activity does not map one-for-one to Mainnet burn
Rollups move much of their execution outside Ethereum Mainnet and publish supporting information in batches. Ethereum's rollup documentation explains the separation between user activity on the rollup and data publication to the base chain. A large number of user transactions can share one publication cost.
That makes transaction count a weak standalone proxy for burn. The useful variables include the resource used, quantity consumed and fee paid per unit. Better compression or more efficient batching can accommodate additional activity without a matching increase in payments to Mainnet.
EIP-4844 introduced blob gas as a separate resource with its own pricing mechanism. The specification also burns the blob fee. Ordinary execution fees and blob fees therefore need separate treatment before they are combined into an ETH-denominated total.
For a hypothetical publication, 200 units of a resource priced at 5 units of fee each produce 1,000 units of fees. If a capacity or efficiency change allows 300 units of use at a price of 2, the result is only 600. More usage and lower aggregate payments are mathematically compatible; actual Ethereum measurements require its real resource units and prices.
Our Layer 2 guide follows execution, verification and exits. The economic lesson here is narrower: adoption, user costs and ETH destruction are connected through an architecture, not a fixed exchange rate between transactions and burn.
Staked or inaccessible ETH is a different category
Depositing ETH into staking does not itself destroy it. The asset remains part of supply while its availability and withdrawal process change. Ethereum's staking overview distinguishes direct participation from pooled and custodial arrangements, which can introduce their own claims and intermediaries.
Likewise, losing a private key can make a balance practically inaccessible without giving the protocol a verified basis to remove it from total supply. A dashboard that excludes an estimated quantity of inaccessible coins is applying a circulating-supply methodology, not necessarily measuring protocol burn.
Wrapping also needs care. An asset receipt backed by ETH held elsewhere can represent the same underlying economic stock. Counting both as separate ETH would inflate the estimate. Readers should identify whether a figure describes native supply, liquid trading availability, a receipt token or an economic claim.
Burn is a mechanism, not a price forecast
A useful dashboard comparison starts by reconciling units. A burn estimate expressed in dollars incorporates an ETH price as well as an ETH quantity. If exactly 100 ETH is destroyed in each of two periods, a higher dollar valuation in the second period does not mean more ETH was removed. Similarly, a cumulative counter keeps rising whenever additional ETH is destroyed even if the pace of destruction is slowing. To compare pace, use the increment over comparable intervals rather than the level of the lifetime counter. The same discipline applies to an annualized estimate: extending a short interval mathematically is not a forecast that its conditions will persist.
Supply accounting can explain what happened to the quantity of ETH. It cannot establish the price investors will pay next. Demand, liquidity, expectations and the terms of financial products remain relevant. A period of net destruction can coexist with falling prices, just as net issuance can coexist with rising prices.
When reading a supply claim, identify the measurement period, separate new rewards from transferred income, distinguish execution burn from blob burn, and check whether the quoted number is total supply or an adjusted circulating estimate. Those checks make the claim interpretable without treating one day's outcome as a permanent property of Ethereum.