EIP-1559 explained: how Ethereum’s base fee and fee burn work
In August 2021 Ethereum stopped running a simple fee auction and started burning part of every fee. Here is the rule, the arithmetic behind it, and the claims about it that go further than the rule does.
Photo: “Point of No Return” by laszlo-photo, CC BY 2.0, via Flickr (edited: cropped and resized).
EIP-1559 replaced Ethereum’s fee auction with a protocol-set base fee per gas, which is burned, plus an optional tip for the validator [1]. The base fee moves by at most 12.5% per block, following demand. It went live with the London upgrade in August 2021 [2].
Key points
- 1Before EIP-1559, users bid a single gas price and paid exactly what they bid; fees were volatile and often overpaid.
- 2Now every block has a base fee set by the protocol from the previous block’s fullness. It is burned, not paid to anyone.
- 3Users add a priority fee (tip) for the validator and a max fee as a ceiling; anything above base fee plus tip is not charged.
- 4Burning fees means ETH has no fixed supply: the EIP’s own authors said they could not predict whether ETH would end up inflationary or deflationary.
- 5EIP-1559 made fees more predictable, not necessarily cheaper.
On this page
- What problem was EIP-1559 trying to solve?
- What did EIP-1559 change?
- How does the base fee move from block to block?
- Why is the base fee burned instead of paid to validators?
- How much ETH does the burn remove?
- Does EIP-1559 make ETH deflationary?
- Did EIP-1559 make gas fees cheaper?
- When did it happen, and what came after?
- What mistakes do beginners make here?
- Frequently asked questions
- The bottom line
- Sources
What problem was EIP-1559 trying to solve?#
An EIP (Ethereum Improvement Proposal) is a public design document for a change to Ethereum. EIP-1559, created on 13 April 2019 by Vitalik Buterin and co-authors, targeted the way transaction fees were priced [1]. Until then Ethereum used a simple auction: each user offered a gas price, block producers picked the highest bids, and everyone paid what they had bid [1].
The EIP’s authors listed several problems with that auction [1]:
- Volatile prices. When blocks were full, bids swung wildly — the EIP contrasts paying 1 and 10 nanoeth per gas — even though one extra transaction cost the network about the same either way.
- Needless delays. A hard per-block limit meant transactions often waited several blocks during short bursts of demand.
- Overpaying. In a first-price auction people must guess what others will bid. Fee estimators often got it wrong, so users frequently paid too much.
What did EIP-1559 change?#
Fee market before and after EIP-1559
Each block now has a base fee per gas, which the protocol calculates from the block before it. The block also gets a gas target, equal to the block gas limit divided by an “elasticity multiplier” of 2 [1]. Blocks can therefore be up to twice the target size when demand spikes, and the base fee rises until demand falls back. Transactions carry two new fields: a max priority fee (the tip) and a max fee, the most the sender will pay per gas in total [1].
| Part | Set by | Goes to |
|---|---|---|
| Base fee per gas | Protocol (from the previous block) | Burned — no one receives it |
| Priority fee per gas | Sender, up to their max tip | The block’s validator |
| Max fee minus (base fee + tip) | Sender, as a ceiling | Never charged; stays with the sender |
In the reference code, the tip actually paid is the smaller of the sender’s max tip and (max fee − base fee), because the base fee is filled first.
How does the base fee move from block to block?#
If the previous block used exactly its target, the base fee stays the same. If it used more, the base fee rises; if less, it falls. The size of the change is the gap from the target, as a share of the target, divided by 8 [1]. Since a block can be at most twice the target, the biggest possible move is one-eighth, or 12.5%, either way [3].
New base fee = Old base fee × (1 + (Gas used − Target) ÷ Target ÷ 8)Simplified from the EIP-1559 reference code, which uses whole-number division and forces a rise of at least 1 wei whenever a block is over target.
| Step | Value |
|---|---|
| Empty block (0% of limit): 20 × (1 − 1/8) | 17.5 gwei |
| Exactly at target (50% of limit) | 20 gwei |
| 75% of limit: 20 × (1 + 0.5/8) | 21.25 gwei |
| Completely full (100% of limit): 20 × (1 + 1/8) | 22.5 gwei |
A run of full blocks compounds: ten full blocks in a row would multiply the base fee by 1.125 ten times — about 3.2 times the starting level. That makes long stretches of full blocks expensive, which is the point [3].
Why is the base fee burned instead of paid to validators?#
The EIP gives several reasons [1]. If block producers received the base fee, they would have an incentive to manipulate it to extract more fees from users. Burning it removes that incentive. Burning also ensures that only ETH can be used to pay for transactions on Ethereum, and it offsets some of the new ETH issued to block producers.
“Burned” has a precise meaning here. In the reference code, the sender is charged base fee plus tip for each unit of gas used, the block producer is credited only the tip, and the base fee is credited to no account at all [1]. That ETH simply stops existing.
| Step | Value |
|---|---|
| Burned = 21,000 × 10 gwei | 0.00021 ETH |
| To the validator = 21,000 × 2 gwei | 0.000042 ETH |
| Total fee = 21,000 × 12 gwei | 0.000252 ETH |
Here five-sixths of the fee is burned. The split changes with the tip you choose and the base fee at the time.
How much ETH does the burn remove?#
It depends on how busy the network is. ETH supply has two opposing forces: issuance to validators, and the burn [2]. ethereum.org gives an illustration from the time of The Merge, when roughly 14 million ETH was staked and validators were issued about 1,700 ETH per day. It asks: what average base fee would burn the same amount?
| Step | Value |
|---|---|
| Blocks per day = 86,400 ÷ 12 | 7,200 |
| Gas per day = 7,200 × 15,000,000 | 108 billion gas |
| Burn at 16 gwei = 108 billion × 16 ÷ 10⁹ | 1,728 ETH per day |
| Issuance it offsets (with ~14 million ETH staked) | about 1,700 ETH per day |
So, in that illustration, an average base fee of about 16 gwei for a whole day would cancel out that day’s issuance. Both inputs move: issuance changes with the amount of ETH staked, and ethereum.org’s blocks page now gives a target of 30 million gas per block [4]. Treat the 16 gwei figure as a worked example of the method, not as today’s threshold.
Validator penalties and slashing also remove ETH without paying it to anyone, so they act as a small extra burn [2]. Since EIP-4844 went live with Dencun [5], scheduled for March 2024 [6], rollups pay a separate blob fee for posting data, and that fee is burned too [7].
Does EIP-1559 make ETH deflationary?#
Sometimes, and nobody can say in advance how often. The EIP’s security section is blunt: by burning the base fee, Ethereum can no longer guarantee a fixed supply. If more is burned than issued, ETH is deflationary over that period; if more is issued than burned, it is inflationary. Because demand for block space cannot be controlled, the authors could not say which it would be [1].
Did EIP-1559 make gas fees cheaper?#
Not by itself. Its goal was predictability: with a known, slowly moving base fee, wallets can set fees reliably, and most users should not have to adjust them by hand even when the network is busy [1]. When many people want block space, the base fee still rises, and tips still decide who gets in first [3]. Lower fees for everyday activity are the job of layer 2 rollups.
When did it happen, and what came after?#
- 13 April 2019EIP-1559 is created as a draft proposal [1].
- August 2021The London upgrade activates EIP-1559 and fee burning begins [2].
- September 2022The Merge moves Ethereum to proof-of-stake; the burn is unchanged and tips now go to validators instead of miners [2].
- 13 March 2024The Dencun upgrade goes live on mainnet; the Ethereum Foundation had scheduled activation for 13:55 UTC [6] and later described blobs as live since Dencun [5]. Its EIP-4844 [6] adds blob transactions with their own base fee, also burned, using a mechanism similar to EIP-1559 [7].
What mistakes do beginners make here?#
- Thinking the max fee is what you pay
The max fee is a ceiling. You pay the block’s base fee plus your tip, and the rest of the cap is never charged.
- Thinking validators earn the base fee
Validators receive only the tip (and, separately, protocol rewards). The base fee is destroyed.
- Calling ETH “deflationary” as a permanent fact
Whether supply shrinks or grows depends on activity over a given period. The EIP’s authors explicitly declined to predict it.
- Expecting EIP-1559 to lower fees
It changed how fees are set, not how much block space exists. Busy periods are still expensive.
Frequently asked questions#
Is EIP-1559 the same as the London upgrade?
EIP-1559 is one proposal; London was the network upgrade, in August 2021, that switched it on [2].
Do old-style transactions still work?
Yes. Legacy transactions are still accepted; their single gas price is treated as both the max fee and the max tip, so they do not benefit from the new pricing [1].
Where does burned ETH go?
Nowhere. The protocol deducts it from the sender and credits it to no account, so it leaves the supply permanently [1].
Why 12.5%?
The EIP divides the base-fee change by a constant of 8, and a block can be at most twice the target, so the largest move is 1/8, or 12.5%, per block [1].
Are blob fees burned too?
Yes. EIP-4844 deducts the blob fee from the sender before execution and burns it; it is not refunded if the transaction fails [7].
The bottom line#
EIP-1559 replaced guesswork with a rule: a base fee that tracks demand block by block and is burned, plus a tip for the validator and a cap you control. It made fees predictable and tied ETH’s supply to how much the network is used.
What it did not do is make fees cheap or ETH’s supply certain. For the day-to-day side of fees, read Ethereum gas fees explained; for the bigger picture, start at what is Ethereum.
Sources#
Grade A = primary source (regulator, protocol specification, client code, original author). Grade B = expert secondary source used for explanation only.
- AEthereum Improvement Proposals. EIP-1559: Fee market change for ETH 1.0 chain, 2019.
- Aethereum.org. How The Merge impacted ETH supply, 2026.
- Aethereum.org. Ethereum gas and fees: technical overview, 2026.
- Aethereum.org. Blocks, 2026.
- AEthereum Foundation Blog (Protocol Support Team). Pectra Mainnet Announcement, 2025.
- AEthereum Foundation Blog (Protocol Support Team). Dencun Mainnet Announcement, 2024.
- AEthereum Improvement Proposals. EIP-4844: Shard Blob Transactions, 2022.
- AEuropean Supervisory Authorities (EBA, ESMA, EIOPA). EU financial regulators warn consumers on the risks of crypto-assets, 2022.
Keep reading
Ethereum gas fees explained: what you pay and why
Every action on Ethereum costs gas. This guide breaks the fee into its parts, works through real arithmetic, and explains why a failed transaction can still cost you money.
What is Ethereum? A plain-English guide to the network and ETH
Ethereum is a public blockchain that runs programs as well as payments. Here is what that means, how a transaction moves through the network, and what nobody can promise you about it.
Ethereum staking explained: how validators earn, and what they can lose
Staking is how Ethereum chooses who adds blocks. Here is what a validator does, how rewards and penalties are calculated, how the different ways to stake compare, and why no staking return is promised.