A blockchain bridge is a tool that connects two separate blockchains so assets and data can move between them. It exists to make different networks talk to each other, which people often call interoperability.
Blockchains usually run with their own rules and cannot natively exchange information with other chains. This isolation limits what users and apps can do across ecosystems. Bridges reduce that gap, letting tokens and messages pass from one network to another and boosting connectivity for things like DeFi.
Most bridges do not literally send coins across a boundary. Instead, they lock the original tokens on the source chain and issue a representation on the destination chain, often called a wrapped token. When you move back, the wrapped version is redeemed and the locked originals are released.
Some bridges depend on a group of known operators to approve transfers. These are often called trusted or federated bridges. Others lean on smart contracts and on-chain verification, which people describe as trustless designs. There are also relay and sidechain styles that pass messages between networks or connect to a dedicated secondary chain. Wrapped-token bridges are common for simple asset transfers.
Beyond coins and tokens, bridges can carry data that lets smart contracts on one chain react to events on another. That enables cross-chain app features and a smoother experience for users who want to move value or instructions between networks.
Bridges are attractive targets. Bugs in smart contracts, weak validator setups, or poor operations can lead to large losses. Even designs that spread trust can face issues like throughput limits or congestion on the underlying chains, which creates bottlenecks for users. Researching how a bridge is secured and governed helps you understand the trade offs.
Check which chains and tokens it supports, how transfers are verified, and whether the contracts or operators have been audited or widely used. Look for clear documentation on fees, wait times, and recovery steps in case something goes wrong. These points reflect the differences between trusted and trustless models and the practical limits described above.