MEV, maximal extractable value, is the profit that can be made by including, excluding or reordering transactions when a block is built, on top of the normal block reward and gas fees (ethereum.org). For an ordinary user the form that matters is the sandwich attack: a bot places a buy before your swap and a sell after it, and you pay the difference through a worse price. On the real Uniswap v2 WETH/USDC pool of November 21, 2026, a 100,000 USDC buy with 1% slippage could be sandwiched for 0.5873 WETH, about $985.

Key takeaways
  • The name: the term comes from the 2019 paper Flash Boys 2.0 as "miner extractable value"; after The Merge it became maximal extractable value.
  • Three main types: arbitrage between DEXs, liquidations in lending markets, and sandwich attacks on swaps; only the last one takes value directly from your trade.
  • Slippage is the budget: a sandwich can take up to the slippage tolerance you set, so 1% on a large swap is a real amount.
  • Protection exists: private RPCs such as Flashbots Protect and MEV Blocker keep a swap out of the public mempool; MEV Blocker reports that bots extracted more than $1.43 billion from Ethereum users.

What MEV Is

MEV is the value that whoever orders the transactions of a block can capture by choosing that order. On Ethereum, a transaction first waits in the public mempool, where anyone can read it. Specialized bots, called searchers, scan the mempool for trades they can profit from and submit their own transactions with instructions on where to place them.

The term comes from the paper Flash Boys 2.0 by Philip Daian and co-authors, submitted on April 10, 2019, which described bots that “frontrun, i.e., anticipate and exploit, ordinary users’ DEX trades” and called the result miner extractable value. Under proof of work, miners chose the order. Since The Merge, validators propose blocks, and ethereum.org now uses the name maximal extractable value.

The Main Types of MEV

ethereum.org lists DEX arbitrage, liquidations, sandwich trading, NFT MEV and a long tail of smaller strategies. The three that touch WETH most often:

Type What the bot does Effect on users
DEX arbitrage Buys a token on the cheaper DEX and sells it on the more expensive one in one transaction Aligns prices across DEXs; no direct loss to a user
Liquidation Repays part of an undercollateralized loan and takes collateral at a discount The borrower pays the liquidation bonus, 5% for WETH on Aave V3
Sandwich Buys before a user’s swap and sells right after it The user gets a worse price, up to the slippage limit

WETH sits in most of these trades because it is the ERC-20 form of ETH that DEX pools and lending markets use (why WETH exists). Liquidations of WETH collateral, and why they happen, are covered in what is Aave.

How a Sandwich Attack Works

A sandwich attack has four steps, as ethereum.org describes them:

  1. The bot sees a large swap waiting in the public mempool, for example a buy of WETH with USDC.
  2. It calculates how much the swap will move the pool price.
  3. It buys WETH just before the user’s swap, at the price before the impact.
  4. It sells WETH just after the user’s swap, at the higher price the user’s swap created.

The user’s swap still goes through, because the price stayed inside the slippage tolerance the user set: the minimum amount the swap accepts. The bot sizes its first buy so that the user receives exactly that minimum. Everything between the expected amount and the minimum becomes the bot’s profit and the pool’s fees.

A Sandwich on the Real WETH/USDC Pool

The example uses the reserves of the Uniswap v2 USDC/WETH pair on Ethereum at block 26,466,837 on November 21, 2026: 8,318,090.44 USDC and 4,958.58 WETH, with the 0.3% swap fee. A user buys WETH for 100,000 USDC with 1% slippage tolerance.

Step Value
WETH the user expects without an attack 58.7292 WETH (average $1,702.73)
Minimum the swap accepts at 1% slippage 58.1419 WETH
Bot buys first 42,218.35 USDC for 24.9654 WETH
User’s swap fills at the minimum 58.1419 WETH (average $1,719.93)
Bot sells right after 24.9654 WETH for 42,972.55 USDC
User receives less 0.5873 WETH, about $985
Bot’s profit before gas $754.20

Calculated with the constant-product formula on the pair’s reserves at block 26,466,837, November 21, 2026; gas and the bot’s payment to the block builder are not included.

The user’s loss is larger than the bot’s profit because part of it goes to the pool as swap fees on the bot’s two trades. With 0.5% slippage the same attack would be smaller; with 3% it would be much larger.

Who Orders the Block: Builders and MEV-Boost

On Ethereum today a block is often not assembled by the validator that proposes it. Under proposer-builder separation, specialized block builders order transactions and bid for the right to have their block used, and the proposer picks the block that pays the highest fee (ethereum.org). MEV-Boost is the software that connects validators to builders through relays, which hold a builder’s block until the proposer signs its header.

For a user this means the order of transactions is decided by builders who compete on how much value they can extract, which is why a swap visible in the public mempool can be sandwiched within the same block.

How to Protect a WETH Swap

No method removes MEV completely, but several reduce the risk of a sandwich to near zero for normal swaps:

  • Set a low slippage tolerance: the slippage tolerance is the most a sandwich can take; 0.5% instead of 3% cuts the maximum loss by a factor of six. Too low, and the swap may fail when the price moves on its own.
  • Use a private RPC: Flashbots Protect sends transactions to a private mempool “hidden from frontrunning and sandwich bots” (Flashbots docs), RPC rpc.flashbots.net/fast. MEV Blocker, run by CoW Protocol, Agnostic Relay and Beaver Build, uses https://rpc.mevblocker.io and returns 90% of any backrun profit to the sender.
  • Split large swaps: price impact grows faster than trade size, so several smaller swaps leave less room for a sandwich than one large swap in a thin pool.
  • Wrap and unwrap are safe: turning ETH into WETH and back through the WETH9 contract is always 1:1 and has no price to move, so there is nothing to sandwich (how wrapping works).

In a self-custody wallet such as WETH Wallet, the swap screen shows the amount you receive before you confirm; checking it before a large swap is the simplest protection.

Limits and Sources

The figures on this page come from different kinds of sources, and some cannot be checked independently:

  • The sandwich example is a calculation, not a recorded attack: it applies the Uniswap v2 formula to real reserves at one block.
  • $1.43 billion extracted is MEV Blocker’s own figure on its homepage, with no date stated; it is a vendor estimate.
  • Protection is not total: MEV Blocker says it prevents 99% of sandwiches and that “no existing solution can provide full 100% protection”.
  • Definitions follow ethereum.org’s MEV page (updated October 28, 2026) and the Flash Boys 2.0 abstract (arXiv 1904.05234).

MEV FAQ

Maximal extractable value. It was first called miner extractable value in the 2019 paper Flash Boys 2.0; after The Merge validators replaced miners, and ethereum.org now uses maximal extractable value.