Bridging crypto between networks: how it works and risks

Three designs, three different risks

Lock and mint. The canonical bridges of most layer 2s work this way: the original token is locked on the source chain and a representation is minted on the destination. Arbitrum documents exactly this for USDC arriving through its bridge, producing the bridged token rather than the issuer’s native one. Your claim on the original depends on the bridge contract continuing to work as intended.

Burn and mint by the issuer. Circle’s Cross-Chain Transfer Protocol burns USDC on the source chain and mints native USDC on the destination. There is no bridge-held reserve and no new representation: you end up with the issuer’s own token on the far side.

Liquidity routes. A third-party bridge or aggregator may simply pay you out of a pool it maintains on the destination chain. This is usually the fastest option. What you are trusting is the operator and the pool, not a lock-up.

Ethereum’s own documentation on bridges is blunt about the general point: bridging introduces trust assumptions that holding on a single chain does not.

Check what you will receive, not just what you send

The output token matters as much as the amount. A canonical bridge may hand you a bridged representation that some exchanges will not credit, while an issuer route hands you the native token that they will. That difference can cost far more than the bridging fee.

Before confirming, read the destination asset name and, where the interface shows it, the contract address. If the result is a bridged version, check whether your onward destination accepts it. Our USDC vs USDC.e guide covers what that distinction changes.

Timing: fast routes and slow routes

Moving to a layer 2 through its canonical bridge is generally quick. Moving back to Ethereum through the same bridge involves a challenge period measured in days on optimistic rollups such as Arbitrum One, Base and OP Mainnet.

Third-party routes that quote minutes for the same journey are doing something different: paying you from destination-side liquidity and settling their own position later. That is a legitimate service, not a trick, but you are paying a fee for someone else to carry the delay and taking their operational risk instead.

Exchange routes are a third option. Depositing on one network and withdrawing on another moves value across chains without any bridge contract, at the cost of going through a custodial account with its own fees and requirements.

Checks before you press confirm

Reach the bridge through a link you already trust, not a search result or a message. Bridge front-ends are a standard phishing target, and the interface can look perfect while the contract underneath is not the one you think.

Then check, in order: the source and destination networks are the ones you intend; the asset arriving is the one you want, not a representation you cannot use; the fee and the estimated time are acceptable; you hold enough of the destination network’s gas asset to do anything with the funds when they land; and, for a first use of any route, send a small test amount.

A bridge that asks for a token approval is asking for permission to move that token on your behalf. Grant what the operation needs and review approvals afterwards — see token approvals and revoking.

When not to bridge at all

If the goal is to get funds to an exchange, check whether it already accepts deposits on the network you are holding. Bridging first and depositing second adds a step, a fee and a risk for no gain.

If the amount is small, bridging fees plus onward gas can consume a meaningful share of it. And if you cannot tell what mechanism a bridge uses or what token you will receive, that is itself a reason to stop: an unclear route is one you cannot check.