What is slippage?

Slippage is the gap between the price a trader expects and the price they actually pay. In traditional markets it is usually small — a few basis points from an HFT firm front-running a retail order. On-chain, it can be dramatic, and it comes from two different sources: the mechanics of automated market makers and the behavior of other actors in the mempool. Either one can turn a reasonable-looking swap into a bad trade.

Mechanical slippage

AMMs quote prices from a formula. For a constant-product pool like Uniswap v2, every trade moves the pool's price by an amount proportional to the trade's size relative to the pool's depth. A trade that is one percent of the pool's depth moves the price by roughly one percent. A trade that is ten percent of the pool's depth moves the price by well over ten percent because the curve gets worse as you go.

This is slippage in the pure mechanical sense: the price you receive is not the quoted spot price, because your own trade is part of the reason the price changed. Deep pools minimize it. Thin pools exaggerate it. A 10,000-dollar swap on the USDC/ETH main pool hardly moves the price at all. A 10,000-dollar swap on a long-tail memecoin with 50,000 dollars of liquidity moves the price massively — and the trader eats the difference.

Concentrated-liquidity pools like Uniswap v3 partially fix this by concentrating capital in the active price range, which gives better execution for in-range trades. It does nothing if the trade is large enough to punch through the range.

Slippage tolerance

Wallet interfaces let users set a "slippage tolerance" — a percentage gap between the quoted price and the worst price the user will accept. If the final execution price is outside that range, the trade reverts. Default tolerances are usually 0.5 percent on major pairs, 1 to 3 percent on smaller tokens.

Low tolerance protects against bad execution but causes transactions to revert often during volatile periods. High tolerance ensures the trade goes through but invites the second kind of slippage: MEV.

MEV and sandwich attacks

See the MEV explainer for detail. The short version: the public mempool is visible to anyone watching. Bots scan pending trades looking for ones with generous slippage tolerances. When they find one — say, a user trying to buy 100,000 dollars of a mid-cap token with three-percent slippage allowed — they can execute a sandwich. Front-run: buy the token before the user's trade, pushing the price up. User's trade: executes at the now-worse price. Back-run: sell the token after, capturing the spread.

The user pays the full slippage they allowed. The bot keeps the difference. Sandwich attacks have extracted billions of dollars from Ethereum users over the years, with the bulk going to a small number of sophisticated searchers.

Private mempools (Flashbots Protect, MEV Blocker) and MEV-aware aggregators (CoW Swap, 1inch Fusion) mitigate this by routing the transaction so that searchers cannot see and sandwich it. For large trades, using these is a material improvement over broadcasting to the public mempool.

Price impact vs. slippage

The terms overlap but are not identical. Price impact is the theoretical, mechanical move caused by the size of the trade itself. Slippage is the actual realized gap, which includes price impact plus any intermediate price movement between the quote and the execution — including MEV, general market movement, and other trades landing in the same block.

A wallet showing "price impact: 2%" is telling you the mechanical component. If you set a slippage tolerance of 3 percent and the trade executes at 2.9 percent worse than quoted, the extra 0.9 percent is the non-mechanical part — other activity and possibly MEV.

Aggregators

Trade aggregators — 1inch, 0x (Matcha), Paraswap, OpenOcean, CoW Swap — are designed in part to minimize slippage. They split trades across multiple pools and venues, routing pieces through whichever paths give the best combined execution. For a large trade, the aggregator might route 30 percent through Uniswap, 40 percent through Curve, 30 percent through Balancer — each chunk small enough not to blow out any single pool.

CoW Swap goes further, using batch auctions and peer-to-peer settlement where possible. For orders that match against other orders in the same batch, slippage is effectively zero because no pool is touched. The aggregator has become a necessary layer for anyone trading seriously on-chain.

Practical guidance

A few rules of thumb. Never set slippage tolerance blindly high; 5 to 10 percent on a mid-cap trade is an invitation to be sandwiched. Check the price impact before signing. For sizable trades, use an aggregator, not the direct DEX front-end. For very large trades, split across blocks or use RFQ-style execution. If a pool's depth is smaller than your trade, walk away and find a different pool or venue.

Why it matters

Slippage is a cost, and it is one of the costs users most consistently underestimate. Gas fees are visible; slippage is often buried three screens deep. Over the course of a year, slippage losses to mechanical impact plus MEV can dwarf gas costs, especially for active traders. Understanding where slippage comes from — and using the right tools to control it — is the difference between trading on-chain efficiently and paying a silent tax on every transaction.