EIP-1559 is the 2021 upgrade that replaced Ethereum's chaotic fee auction with a predictable base fee that the network sets automatically and then destroys. The change made fees easier to estimate and turned a slice of every transaction into permanently removed ETH. This guide explains what it changed, why the burn exists, and whether it really makes ETH scarce.
What is EIP-1559?
EIP-1559 is an Ethereum upgrade, live since August 2021, that rebuilt how transaction fees are priced and introduced a burn that destroys part of every fee. It split the fee into an automatic base fee set by the network and an optional tip set by the user, then made the base fee disappear instead of paying it to anyone. It is one of the most consequential changes Ethereum has ever shipped.
The upgrade arrived inside a network update called the London hard fork on August 5, 2021 (source: EIP-1559 specification). Before it, paying an Ethereum fee felt like a guessing game; after it, the network publishes a clear per-block price you can read in advance. The headline mechanism is the part people remember: the base fee is not collected by anyone, it is burned, meaning the ETH is destroyed and gone from the supply forever.
A simple way to picture the shift is a tollbooth that posts its price instead of running a silent auction. The old booth made every driver shout a bid and hope; the new booth shows a number on a screen that rises when traffic is heavy and falls when it clears. You can still pay a little extra to jump the queue, but you always know the baseline. That posted, self-adjusting price is the heart of EIP-1559, and the rest of this guide unpacks why it was built that way and what the burn does to ETH over time.
How did Ethereum fees work before EIP-1559?
Before EIP-1559, Ethereum ran a first-price auction where every user blindly bid a fee and hoped it was high enough to get included. There was no posted price, so you guessed, and guessing wrong meant either overpaying badly or having your transaction stall. During busy periods this turned into an expensive scramble.
The mechanics made the problem worse, and a quick picture shows why. Each user attached a gas price to their transaction, and block producers simply picked the highest bidders, so it worked like a sealed-envelope auction where nobody could see what others offered. To be safe, people overbid, padding their fee far above what was actually needed just to avoid getting stuck (source: Coinbase: what is EIP-1559). When demand spiked, a popular token launch or a market panic, the bidding war exploded and fees became wildly unpredictable, sometimes swinging several times over within minutes. Worse, a losing bid could still cost you. A transaction that failed to get in wasted time and sometimes gas, so users routinely paid for a certainty they did not really get. The old fee market is, in spirit, the same kind of pure auction that still drives some other networks' busy-period costs, as our guide on why Bitcoin fees spike shows.
There was a second, quieter problem: the entire fee went to the miner. That gave block producers an incentive to keep blocks artificially full and fees high, which is the opposite of what users want. EIP-1559 was designed to fix both issues at once, the guesswork and the misaligned incentive, by posting a price and then taking most of it out of any single party's pocket. How it did that is the next section.
What actually changed after EIP-1559?
After EIP-1559, the fee split into a predictable base fee that the protocol sets and burns, plus an optional tip that goes to the validator. The base fee adjusts automatically block by block, rising when blocks are full and falling when they empty, so you no longer have to guess the market price. You only decide how big a tip to add for speed.
The before-and-after contrast is the clearest way to see the upgrade (source: OKX: EIP-1559 explained).
Feature | Before EIP-1559 | After EIP-1559 |
|---|---|---|
How the price is set | Blind first-price auction | Automatic base fee, posted per block |
Predictability | Low, frequent overbidding | High, you can read the base fee in advance |
Where the fee goes | All to the miner | Base fee burned, tip to the validator |
Your main decision | Guess a winning bid | Choose a tip for priority |
The mechanism that drives this is a simple feedback loop. Each block has a target size, roughly half of its maximum, and if blocks run fuller than that target, the base fee ticks up for the next block; if they run emptier, it ticks down. The change is capped at about 12.5 percent per block, so the fee can only climb so fast, which is what makes it predictable: a sudden rush still takes several blocks to push the fee up sharply (source: Ethereum.org gas documentation). That predictability comes with a trade-off worth naming. Blocks are now elastic, able to expand up to double the target when demand spikes, so EIP-1559 traded steadier fees for more variable block sizes.
There is also a change you see directly in your wallet. Instead of one gas-price field, you now set two: a max fee (the most you will pay per unit) and a max priority fee (your tip). The tip became the small lever you control: a normal tip during calm periods, a larger one when you genuinely need the next block, and you are refunded the difference between the max fee and what the block actually charged. A small example makes it concrete: if the base fee is 10 gwei and you add a 1 gwei tip, you pay 11 gwei for every unit of gas, the 10 gets burned, and the 1 goes to the validator who includes you.
Your gas fee Base feeburned, removed forever Priority tippaid to the validator
Does any of this affect you directly? If you only trade on an exchange, barely, because no on-chain transaction happens. EIP-1559 shows up the moment you move crypto on-chain, such as withdrawing to your own wallet, where you pay the base fee plus your tip. Importantly, EIP-1559 made fees more predictable, not automatically cheaper, the dramatic price drops of 2026 came mostly from Layer-2 networks, not from this upgrade alone.
Why does Ethereum burn the base fee?
Ethereum burns the base fee to remove the incentive for block producers to manipulate fees, and as a side effect, it ties ETH's supply to how heavily the network is used. If the base fee were paid to validators, they would benefit from keeping fees high; burning it means no one profits from congestion. The burn turns a fee into a supply event.
The design solves the incentive problem cleanly, and the logic is worth following. Because the base fee is destroyed rather than collected, a validator gains nothing by stuffing blocks to inflate it, so the price can reflect real demand instead of being gamed. That leaves the tip as the only part a validator actually earns, which keeps their incentive pointed at including transactions quickly rather than congesting the chain. From Blofin's operational perspective, a common point of confusion is users assuming the burned base fee is some charge an exchange or middleman pockets; it is not, the base fee is destroyed by protocol rule and reaches no company at all.
The side effect is the part that captured imaginations. Every burned base fee permanently shrinks the ETH supply by that amount, so the more the network is used, the more ETH is destroyed. Since the upgrade, the burn has removed a large amount of ETH from circulation, roughly 4.6 million ETH as of 2026, though the figure keeps climbing and is best checked on a live dashboard (source: Ultrasound.money burn tracker). That linkage between usage and scarcity is what fuels the "deflationary" talk, which deserves an honest look.
Did the burn actually make ETH deflationary?
Not permanently, and this is where careful wording matters: the burn can make ETH deflationary, but only when it destroys more ETH than the network creates. Ethereum still issues new ETH to pay validators, so whether the supply shrinks or grows on any given day is a tug-of-war between issuance and the burn. Heavy usage can tip it into shrinking; quiet usage lets it grow.
The honest 2026 picture is more nuanced than the slogans suggest. In busy stretches after Ethereum switched to proof of stake, the burn at times outpaced new issuance, and the total ETH supply genuinely fell for a while, which is what gave rise to the "ultrasound money" idea. But fees in 2026 are very low, so the burn is small, and new issuance to validators currently outpaces it, leaving ETH mildly inflationary rather than deflationary. Both states are possible by design, and which one holds depends entirely on network activity, a point our companion guides on Ethereum tokenomics and the ultrasound-money thesis examine in full. Treating ETH as permanently deflationary is simply inaccurate.
The cleaner takeaway is structural, not directional. EIP-1559 did not promise that ETH would always shrink; it built a mechanism where usage directly removes supply, so the asset becomes scarcer precisely when demand for block space is high. Compare that with Bitcoin's approach of a fixed cap and scheduled halvings, a contrast our guide on Bitcoin versus Ethereum draws out. Neither design is automatically better. EIP-1559 simply chose to link scarcity to real use rather than to a fixed schedule.
Frequently asked questions
Did EIP-1559 make Ethereum fees cheaper?
Not directly; it made them more predictable. The upgrade replaced blind bidding with a posted base fee that adjusts automatically, which removed wild overbidding but did not by itself lower the underlying cost of block space. During congestion, the base fee still rises. The large fee reductions Ethereum users enjoy in 2026 came mainly from Layer-2 networks and later upgrades, not from EIP-1559 alone. Its gift was clarity and a fairer fee market, not cheapness on its own.
Where does the burned ETH actually go?
Nowhere; it is permanently destroyed. When the base fee is burned, that ETH is removed from the total supply and cannot be recovered or spent by anyone, including the network's developers. There is no wallet that collects it and no company that benefits. You can watch the destruction happen in real time on public burn dashboards, which tally how much ETH each block removes. The effect is simply a smaller supply than there would otherwise be.
Can I avoid paying the base fee?
No, the base fee is mandatory for any mainnet transaction, though its size depends on congestion. You cannot opt out of it the way you can choose your tip, because it is the protocol's automatic price for using block space. What you can do is reduce it indirectly. Transacting during quiet hours keeps the base fee low, a timing habit our guide on how to choose Bitcoin fees applies on another chain. Using a Layer-2 network drops the equivalent cost to a tiny fraction of mainnet. The base fee is a floor you manage around, not one you skip.
Is ETH inflationary or deflationary right now?
As of 2026, ETH is mildly inflationary, because low fees mean the burn destroys less ETH than the network issues to validators. This can flip: during periods of heavy network use, the burn has at times exceeded issuance and the supply shrank. The honest answer is that ETH's supply direction is not fixed, it swings with activity, so any claim that ETH is permanently deflationary is wrong. A live supply dashboard shows the current state better than any static figure.
Who proposed EIP-1559 and why?
It was proposed by Ethereum researchers, including co-founder Vitalik Buterin, to fix two problems: unpredictable fees and the incentive for miners to keep fees high. The proposal went through years of public debate before shipping, and it drew strong opposition from miners who stood to lose the full fee revenue the old system gave them. That tension is a textbook example of Ethereum's open governance, where a change can advance even when a powerful group resists it, provided the broader community and developers support it.
Does the priority tip still matter after EIP-1559?
Yes, the tip is how you buy speed. Since the base fee gets your transaction onto the chain at the going rate, the tip is the extra you add to be included sooner, which matters most during busy periods. When the network is calm, a small default tip is usually enough and you rarely need to think about it. When blocks are full and you need to act fast, raising the tip moves you up the queue. It is the one fee lever that remains fully in your hands.
Researched and written by the Blofin Academy editorial team with AI-assisted drafting. Primary sources include the EIP-1559 specification and Ethereum.org gas documentation, with burn figures drawn from public on-chain supply dashboards. All facts independently verified against cited sources current as of June 2026.
This article is for informational purposes only and does not constitute financial advice, investment guidance, or a recommendation to buy, sell, or hold any digital asset. ETH issuance, burn, and supply figures are protocol-driven on-chain measures that change frequently and are presented as approximate as of 2026, not forecasts; whether ETH is inflationary or deflationary at any moment depends on network activity. Cryptocurrency markets involve significant risk and you should conduct your own research and consult qualified professionals before making decisions.