A state channel is a scaling technique in which two or more parties lock funds or assets in a smart contract and then interact off the blockchain by exchanging signed messages. Only the opening deposit and the final result, or a dispute, are recorded on-chain. Everything in between happens directly between the participants, almost instantly and without a fee for each step.
State channels are the general form of the payment channel. A payment channel tracks only who owns how much, while a state channel can track any shared "state": the position in a game, the terms of a bet, or the balance of a running service bill. The idea was popularised in a 2015 essay by Ethereum developer Jeff Coleman. It belongs to the family of off-chain scaling methods and is often grouped with Layer 2 networks.
State channels in plain terms
Imagine two friends playing a long game of chess for a small stake. They give the money to a trusted referee at the start, then play the whole game at a café, each signing the score sheet after every move. Only when the game ends do they show the signed sheet to the referee, who pays the winner.
If one friend storms off mid-game and refuses to sign, the other can bring the last signed sheet to the referee. The referee waits a set time for the other side to show a newer sheet, then settles based on the latest one. A state channel works the same way, with a smart contract as the referee and digital signatures in place of handwriting.
How a state channel works
- Opening. Participants deposit funds into a channel contract, often a multi-signature arrangement, in one on-chain transaction.
- Updating. They exchange new versions of the state, each with a higher sequence number and signed with every participant's private key. A public key lets anyone check those signatures later.
- Closing cooperatively. When they are done, all parties sign the final state and submit it, and the contract pays out accordingly.
- Disputing. If someone disappears or submits an old state, the other party can submit a newer signed one during a challenge period. Many designs penalise the cheater by forfeiting their deposit.
Because the chain accepts the latest signed state, each participant needs to watch the blockchain during the challenge window, or hire a third party to watch for them.
Researchers have also proposed channel networks and "virtual channels". In these, two people who share no direct channel can still interact through an intermediary that has channels with both, so they need no new on-chain deposit. The Lightning Network uses a simpler version of this idea to route payments across several hops.
Uses and examples
- Payments. The best-known example is Bitcoin's Lightning Network, a network of payment channels that routes payments across several hops.
- Streaming and metered payments. Channels suit pay-per-use services and microtransactions, where an on-chain fee would cost more than the payment.
- Games and turn-based applications. Each move can be a signed update, with the chain used only to settle the outcome.
- Ethereum projects. Projects such as the Raiden Network (payment channels for tokens) and generalised state-channel frameworks were built on the Ethereum blockchain to cut gas fees for repeat interactions.
State channels compared with other scaling methods
| Approach | Where activity happens | Who can join |
|---|---|---|
| State channel | Off-chain, between fixed participants | Only those who opened the channel |
| Rollup | Off-chain execution, data posted on-chain | Anyone |
| Sidechain | A separate chain with its own validators | Anyone |
Unlike optimistic rollups and sidechains, a channel puts no transaction data on-chain until settlement, which makes it cheap and private. The trade-off is that it serves only a fixed group. Sharding takes a different route, raising the base layer's own capacity.
Limitations
- Locked capital. Funds stay tied up in the channel until it closes, and a channel can move only as much as was deposited.
- Liveness requirement. Participants, or a watching service, must stay online to challenge stale states, which weakens the trustless promise for casual users.
- Fixed membership. Adding a new party usually means opening a new channel, so channels fit repeated interactions between the same people, not open markets.
- Poor composability. State inside a channel cannot be used directly by other on-chain contracts.
- Delayed finality. Off-chain updates are final between the parties, but on-chain settlement finality depends on the closing transaction and any challenge period.
These limits are why general-purpose state channels saw less use on Ethereum than rollups, while payment channels on Bitcoin kept growing. The differences between on-chain and off-chain activity are covered further in On-Chain vs Off-Chain.