Remove Liquidity on Base

Connect the wallet that holds it and everything you own surfaces on its own, whether the liquidity sits as a v2 LP balance or a v3 or v4 position NFT. Pick how much to take out, from 1% to everything, read the payout breakdown, and sign once. The version differences are handled underneath, including the extra call that v3 needs before it will actually pay you anything.

Because the price moved outside your price range, and on Base this is the normal outcome rather than the exception for anyone providing liquidity. Sampling real Uniswap v3 removals on the chain, around 84% of them were out of range at the moment of withdrawal. When that happens the market has already converted your principal into whichever side of the pair it ended on: all quote asset if the price rose through your range, all token if it fell through. Your fees are tracked separately and usually still arrive in both tokens, because they built up over the stretch when the position was still active.

Yes, and this is the part most tools get wrong. On Uniswap v3 the call that reduces your liquidity does not actually send you anything at all. It credits what you are owed inside the position manager and stops there, so a withdrawal built from that call alone moves zero tokens to your wallet, principal included. A second call is what actually pays out. Here both are issued inside one transaction, so what you deposited and what you earned reach your wallet at the same moment. On v4 the protocol handles all of it in one operation.

Yes. It is worth flagging because the tools that will happily open a v4 position for you on Base mostly cannot get you back out of one, and several publish a v4 create page with no matching removal page at all. v4 keeps positions as NFTs like v3 but settles the withdrawal in a single operation rather than two.

Any share from 1% to 99% works, and whatever stays behind carries on earning. Concentrated positions hold their bounds through a partial withdrawal, which means you can shave the same one down repeatedly without minting a replacement or losing the range you originally picked.

The position empties and, on v3 and v4, the NFT is burned. Uniswap's own interface does not do this: it leaves the spent NFT in your wallet, so people often expect to keep it. Burning reclaims the storage and retires that token ID for good. Planning to use those bounds again? Stop at 99% and the position survives, ready to be topped up later. A v2 position has no NFT, so none of this applies; the LP tokens are simply redeemed.

No, and it is straightforward to check for yourself. A pool's price only changes when somebody trades against it, because a swap is the only thing that writes the price slot. Pulling depth out leaves that slot alone. On a real Base pool where one withdrawal took roughly 89% of the active liquidity, the recorded price and tick were byte-for-byte identical in the block before and the block after. What it does change is the pool's sensitivity from then on: with less depth in place, the following trade travels further than it otherwise would.

0.002 ETH plus gas, with no cut of what you withdraw. Be aware which of those two numbers actually matters on this chain: a full v3 withdrawal on Base runs about 230,000 gas, which comes to a fraction of a cent, so the service fee is the cost here and the network is close to free. Other tools in this category publish anywhere from 0.001 to 0.015 ETH for the same action and some publish nothing at all, which puts us second cheapest rather than cheapest. Uniswap's own interface charges nothing beyond gas.

Every Uniswap v2 pair carries a floor of 1,000 LP units that nobody can withdraw, destroyed on the pair's very first deposit as protection against a rounding attack. A pool's creator taking everything back out still leaves that floor sitting there, which is why the closing figure never lands on a round number. Looking at Base pairs in bulk shows how routine this is: in a random sample, about a third had already been emptied down to precisely that untouchable floor and nothing else. Do not try to measure it from the burn address, though. Sending LP tokens to a dead address is something anyone can do voluntarily, and some pairs have nearly their whole supply parked there; only the original 1,000 is enforced by the contract itself.

Be careful, because the obvious answer is the wrong one. USDbC is the bridged dollar Base launched with before Circle issued USDC natively, and it still redeems one for one, but it has shrunk to roughly a seven-hundredth of native USDC's supply. The problem is depth: total USDbC liquidity across Uniswap on Base is now on the order of tens of thousands of dollars, and the direct route into native USDC holds only around twelve thousand. Swapping a meaningful amount there will cost you far more in price impact than the position was ever going to earn. Depositing it to an exchange that credits Base dollars, or bridging it, are the routes that do not eat your balance. Also note that the old bridge.base.org address is retired, so any guide still pointing you there is out of date.

No to both, and that is the point of them rather than a shortcoming in any tool. A lock hands your LP to a contract that refuses to give it back before a date you can read off that contract yourself, and nothing external can shorten it. Burning is the harsher version: the LP went to an address with no owner at all, so what it represented stays where it is for good. If somebody offers to reverse either one for a fee, they are describing an outcome the code does not permit, which makes it a scam rather than a service.

Wait for the sealed block, which takes about two seconds, rather than trusting the first thing your wallet shows you. Base streams sub-second previews of pending state, and a wallet reading that stream can display a success and an updated balance that the final block then contradicts. It is a real quirk of this chain and it does not happen on Ethereum mainnet. Two seconds settles it for practical purposes; batching to Ethereum a few minutes later is what makes it irreversible.