
MEV — maximal extractable value — is the extra value someone can capture by controlling which transactions go into a block and in what order, on top of the ordinary block reward and gas fees. The term replaced “miner extractable value” after Ethereum’s move to proof of stake, since validators now occupy that ordering role instead of miners, but the underlying idea is unchanged: ordering is itself valuable, and someone captures that value.
Why ordering is worth money
Picture two transactions arriving for the same block: a large trade that is about to move a token’s price on a decentralized exchange, and a smaller trade that would benefit from trading just ahead of or just behind it. Whoever decides the order these land in can insert their own transaction to profit from that price movement, or sell that right to someone else who will. Nothing about this requires any transaction to be invalid or any rule to be broken — it’s a direct consequence of the fact that execution order is not fixed by anything other than whoever builds the block.
The roles under proposer-builder separation
Ethereum’s current MEV landscape runs through a structure called proposer-builder separation (PBS), which splits block production into distinct roles rather than leaving one party to do everything:
- Searchers scan pending transactions and on-chain state for profitable opportunities — arbitrage, liquidations, favorable ordering around a large trade — and submit bundles, often with a sealed bid attached, asking to have that bundle included in a specific position.
- Builders are specialized parties that assemble full blocks out of ordinary transactions and searcher bundles, competing to construct the most valuable block they can, then bid for the right to have their block proposed.
- Relays sit between builders and proposers: they check a builder’s block for validity and hold its full contents in escrow, revealing only the header to the proposer until the proposer commits to it, then release the full block for publication. Both builders and proposers have to trust the relay to behave honestly in this role.
- Proposers — validators selected to produce the next block — pick the highest-paying bid among the blocks offered to them and sign a commitment to it, without needing to have built its contents themselves. A proposer can still build a block locally as a fallback if no external bid is attractive.
The commit-reveal structure, where a proposer commits to a block header before seeing its full contents, is a deliberate design choice: it limits how much a builder can be undercut after revealing a profitable opportunity, and it reduces the advantage a very large staking operation would otherwise have if it could simply build every block itself and keep all the MEV in-house.
The block builder auction
Builders compete for block space in something that functions like an auction: each submits a bid representing how much it is willing to pay the proposer for the right to have its block included. The proposer, in turn, almost always takes the highest bid, since that is the most direct way to maximize its own revenue from the slot. This effectively turns block production into a market where validators sell the right to order transactions, and builders compete on how efficiently they can extract and bundle value to win that right.
Common MEV strategies
DEX arbitrage
Buying a token on one exchange and selling it on another within the same transaction, capturing a price difference between them. This is the most heavily competed-over form of MEV, since the opportunity is purely mechanical and visible to anyone running the right software.
Liquidations
Racing to liquidate an under-collateralized loan on a lending protocol and collecting the liquidation fee or discount that protocol offers for doing so. This keeps lending markets solvent, but the race to be first is itself a competitive MEV opportunity.
Sandwich trading
Placing one trade immediately before a large pending user trade and another immediately after it, profiting from the price movement the user’s own trade causes. This is the strategy most directly harmful to ordinary users, since it worsens the price they actually receive compared to what they would have gotten without the sandwich.
Frontrunning
Copying a profitable transaction visible in the public mempool and resubmitting it with a higher fee to get processed first. Private transaction submission channels were built partly as a response to this — by not broadcasting a transaction to the public mempool at all, a user denies would-be frontrunners the visibility they need to act on it.
Is MEV bad?
It depends which strategy and whose perspective. Arbitrage and liquidations generally perform a function markets need — correcting price discrepancies and keeping lending protocols solvent — even though the parties doing it are acting purely out of profit motive, not altruism. Sandwich attacks, by contrast, extract value directly from an ordinary user’s trade with no offsetting benefit to that user. The existence of MEV as a category is a structural fact about any system with public, orderable transactions; which specific strategies show up and how much they cost ordinary users is the part worth paying attention to.
MEV on rollups looks different
Proposer-builder separation is an L1 structure, built around many competing builders and a market-driven relay system. Most rollups today run a single sequencer that performs the equivalent ordering role alone, with none of PBS’s competitive, multi-party checks in place. That does not automatically mean an L2 sequencer behaves badly — but it does mean the structural protection PBS provides on L1, where a proposer choosing to behave badly risks losing out to a competing builder’s bid, has no direct equivalent in most L2 designs today. See the companion guide on sequencer centralisation for how that plays out in practice.
Where the opportunity comes from in the first place
MEV does not appear out of nowhere — it exists because decentralized exchanges, lending markets, and other on-chain systems create predictable, visible price relationships that update only when someone submits a transaction to update them. Anyone watching the public chain state can calculate exactly what a profitable transaction would look like before submitting it, which is precisely what searcher software automates at scale, continuously and faster than a person could by hand.
Flashbots and similar private-bundle infrastructure emerged specifically to manage this: instead of every searcher broadcasting bids publicly and racing on gas price alone — which wastes gas and clutters the network with failed transactions — bundles are submitted privately to builders, who can evaluate and combine them off-chain before only the winning combination ever reaches the public chain.
None of this changes with every new rollup upgrade overnight; it is a slow-moving area of infrastructure design, and the fairest reading of any specific project’s current MEV mitigations is its own technical documentation, not a general claim made about the ecosystem as a whole.
Risk note
MEV is a structural feature of any blockchain with public, orderable transactions, not a bug specific to Ethereum. Understanding it helps explain why trade execution can be worse than expected on a public mempool, and why private submission channels and slippage protection exist; it is not a signal to time trades around.
Not financial advice. Nothing on this page is a recommendation to buy, sell, stake, bridge, or avoid any asset or protocol.
FAQ
Does MEV mean my transaction will be front-run?
Not necessarily. Frontrunning targets visible, profitable opportunities in the public mempool; routine transfers with no exploitable price impact are not attractive targets. Private submission channels further reduce this exposure for sensitive trades.
Who actually keeps the MEV captured in a block?
It is split among the roles involved: searchers who found the opportunity, builders who included it, and proposers who are paid for choosing the winning block, through the bidding process described above.
Is MEV unique to Ethereum?
No. Any blockchain where transaction ordering is decided by a party before inclusion has some form of extractable ordering value; Ethereum’s proposer-builder separation is one specific, widely studied way of managing it.
Can I avoid being sandwiched?
Using slippage limits on trades and private transaction submission channels reduces exposure, though neither eliminates the possibility entirely. This guide explains the mechanism rather than recommending specific tools.
Frequently asked questions
Does MEV mean my transaction will be front-run?
Who actually keeps the MEV captured in a block?
Is MEV unique to Ethereum?
Can I avoid being sandwiched?
Educational content, human-edited and possibly AI-assisted. It describes how systems work and is not financial, legal or tax advice. Report an error via our corrections page.


