MEV — originally “Miner Extractable Value”, now more commonly “Maximal Extractable Value” since Ethereum moved to proof-of-stake — is the profit that can be captured by whoever controls the order of transactions in a block. When you send a transaction to an Ethereum node, it sits in a public mempool before being included in a block, and during that time, anyone watching the mempool can see what you are about to do. A block proposer (or a searcher running bots on their behalf) can then reorder, front-run, or sandwich your transaction in ways that extract value from you. That value is MEV, and it has been one of the defining structural problems of Ethereum’s trading ecosystem since 2019.
The canonical example is a sandwich attack on a DEX trade. You submit a large swap on Uniswap. A searcher sees it in the mempool, calculates that your trade will move the price by 2 percent, and submits two transactions: one to buy the token just before yours (pushing the price up), and one to sell it just after yours (capturing the price impact you created). The searcher profits from the price move that your trade caused. You get worse execution than you would have in an MEV-free world. The difference between what you got and what you should have gotten is MEV extracted from you.
The Rough Shape of the Problem
The scale is not small. Flashbots, the organisation that has done the most work to study and mitigate MEV on Ethereum, estimates that several billion dollars in total MEV has been extracted on Ethereum over the years, with hundreds of millions per year as a steady-state rate. That is not revenue the network sees as spam or hacks — it is a structural tax on on-chain trading that falls primarily on retail users who do not know how to protect themselves.
The mechanisms span a spectrum from relatively benign to outright predatory. Simple arbitrage between DEXes (buying a token on Uniswap and selling on Sushiswap for a slightly higher price) is MEV, and it actually improves market efficiency. Liquidation bots competing to liquidate underwater positions on Aave or Compound is MEV, and it is a necessary service for those protocols to function. Sandwich attacks on naive swaps are MEV, and they are pure extraction with no useful side effect. Generalised front-running, where a bot reads an incoming transaction and copies its strategy to run it first, is MEV at its worst.
Proposer-Builder Separation and MEV-Boost
Ethereum’s response has been architectural rather than trying to eliminate MEV entirely. The current system, built around MEV-Boost, separates the role of “proposing a block” from “building a block”. Validators propose blocks but outsource the construction of the block to specialised builders, who compete by assembling the most profitable block they can from a combination of public mempool transactions and privately-submitted bundles from searchers. Validators pick the highest-bidding block and include it. The searchers capture some of the MEV, the builders capture some, the validators capture some, and users still experience the extraction — but the extraction happens in a more orderly auction rather than through an uncoordinated scramble.
This arrangement has been controversial because it formalises MEV rather than removing it, and because it introduces new concentration: at various points over the last two years, a small number of builders (Flashbots, Titan, Beaverbuild) have accounted for the majority of blocks, which creates a censorship vector. During 2022-2023, Flashbots-built blocks were observed excluding transactions involving OFAC-sanctioned addresses, which raised concerns about protocol-level compliance creeping into Ethereum. The Ethereum community has pushed for “censorship resistance” through multiple builders and “inclusion lists” that let validators force transactions into blocks even if builders want to exclude them, but the dynamic is still evolving.
How to Protect Yourself
Individual users have a few options. Private mempools (Flashbots Protect, MEV Blocker, BloxRoute’s BackRunMe) let you submit transactions directly to block builders without going through the public mempool, which prevents most sandwich attacks. Using a DEX aggregator with MEV protection built in (CoW Swap uses batch auctions that naturally resist sandwiching; 1inch’s Fusion mode does similar work) is another approach. Trading on L2s where the sequencer has a private mempool gets much of the benefit for free — Arbitrum, Optimism, and Base all have significantly less MEV than Ethereum mainnet for the same reason. For large trades, none of these are perfect, but they close most of the gap.
The thing you should not do is submit a large DEX swap through a default wallet with default settings and expect to get a fair price. The ecosystem has moved beyond that being acceptable, and the tools to protect yourself exist — you just have to use them.