EIP-1559

EIP-1559 is an Ethereum Improvement Proposal that changed how transaction fees are set and paid on the Ethereum blockchain. It replaced a simple auction, in which users bid a gas price and miners picked the highest bids, with a protocol-set base fee per unit of gas that rises and falls automatically with demand, plus an optional priority fee, or tip, paid to whoever produces the block. The base fee is not paid to anyone: it is burned, permanently removing that ether from supply. The proposal was activated on 5 August 2021 as part of the London hard fork.

EIP-1559 was first proposed in 2019, with Vitalik Buterin among its authors, building on earlier research into blockchain fee markets. It is one of the most consequential changes in Ethereum's history because it affected user experience, block producers' income and the monetary policy of ether at once. Its base-fee mechanism has since been copied by other blockchains and reused within Ethereum itself, for example for the blob fee market added by the Dencun upgrade.

Before EIP-1559, sending an Ethereum transaction was like bidding at a crowded auction without seeing the other bids. Users guessed how much to offer, often overpaid in busy periods, or waited a long time if they bid too low. EIP-1559 replaced this with something closer to a posted price, like a toll road whose charge goes up when traffic is heavy and down when the road is quiet.

Everyone pays the posted toll, and that money is destroyed rather than given to the road operator, so the operator has no reason to manipulate the price. Drivers who are in a hurry can add a tip to get waved through first. The result is fees that are easier to predict, even if they are not always lower.

Each transaction specifies a maximum total fee per gas and a maximum priority fee. The user pays the base fee plus the tip, and any difference up to the maximum is refunded. Blocks have a target size, half of the maximum. After each block, the base fee adjusts according to how full that block was:

• $b_n$ is the base fee of the current block and $b_{n+1}$ that of the next block.
• $g_n$ is the gas used in the current block.
• $g_{\text{target}}$ is the target gas per block, which was 15 million at launch.

Because a block can hold at most twice the target, the base fee can rise or fall by at most 12.5% per block. For example, if the base fee is 20 gwei and a block is completely full, using 30 million gas against a 15 million target, the next base fee is $20 \times (1 + \frac{1}{8} \times 1) = 22.5$ gwei. If the next block is empty, the fee falls by 12.5%, to about 19.7 gwei. A run of full blocks therefore raises fees quickly until demand eases. Gas limits and targets have since been raised by later network decisions.

Burning the base fee turned network use into a steady drain on the supply of ether. Before September 2022, new ether was issued to miners as block rewards, so burning only slowed supply growth. After Ethereum's switch to proof of stake, issuance fell sharply, and during busy periods the amount burned exceeded new issuance, making supply shrink. This gave rise to the "ultrasound money" slogan among some supporters. The balance depends on activity: when fees fell after layer-2 networks moved much activity off the main chain, supply resumed modest growth. The total destroyed is tracked as burned supply.

Fees became more predictable, and wallets could estimate a sensible price automatically, though EIP-1559 did not by itself make fees cheaper: when demand is high, the base fee climbs. Miners, who previously collected the whole fee, lost the base-fee portion, and some mining pools and groups objected publicly before the London upgrade, but the change went ahead. Today validators receive the tips, along with any extra value from ordering transactions. Pending transactions in the mempool now wait mainly when their maximum fee is below the current base fee.

Critics noted that the mechanism does not raise capacity, so gas fees can still spike during congestion, and that tips and private transaction channels still let sophisticated users pay for priority. Some economists questioned whether burning fees was the best use of the revenue compared with paying it to the network's security providers. Supporters respond that burning prevents block producers from gaming fees and spreads the benefit to all ether holders.

• EIP-1559: Fee market change for ETH 1.0 chain