Bitcoin Multisig Explained: Why Two Keys Beat One
Multisignature wallets require more than one key to move funds, which means no single device, seed, or manufacturer failure can lose your Bitcoin. Here is how they work and when they are worth the extra effort.
Most Bitcoin is held in single-signature wallets: one seed phrase, one set of keys, one point of failure. If that seed is lost, stolen, or — as the July 2026 Coldcard firmware incident demonstrated — generated with insufficient randomness, the coins are gone.
Multisignature wallets remove that single point of failure. Instead of one key controlling the funds, several keys exist and a defined threshold of them must sign before anything moves. This guide covers how multisig works, the common configurations, and whether the added complexity is worth it for you.
How multisig works
A multisig wallet is defined by two numbers: how many signatures are required, and how many keys exist in total. A 2-of-3 wallet has three keys and requires any two of them to authorize a transaction. A 3-of-5 has five keys and requires three.
The rules are enforced by Bitcoin's own scripting system, not by any company or service. When you create the wallet, the participating public keys are combined into a script that specifies the threshold. Spending requires producing valid signatures from enough of the corresponding private keys to satisfy it. No intermediary can override this.
What multisig actually protects against
- Single device compromise. If one hardware wallet is stolen, tampered with, or found to have a firmware flaw, the attacker still cannot move funds with one key alone.
- Single seed loss. Lose one backup in a fire or flood and you can still recover using the remaining keys.
- Vendor risk. Using devices from different manufacturers means no single company's engineering mistake can reach all of your keys.
- Coercion. Keys stored in geographically separate locations mean no one person in one place can be forced to hand over spending authority.
- Inheritance. Multisig makes it possible to design estate arrangements where heirs or a lawyer hold keys that only become useful in combination.
The common configurations
2-of-3 is the standard for individual holders. Three keys on three separate devices, ideally from different manufacturers, stored in different locations. Any two can spend. One can be lost entirely without consequence. This is the configuration most self-custody guidance converges on for meaningful holdings.
3-of-5 suits larger holdings or shared control among several people. More redundancy, more operational overhead.
2-of-2 is generally a bad idea for individuals. It removes the redundancy that makes multisig valuable — lose either key and the funds are permanently frozen.
The vendor-diversity principle
The Coldcard incident made a previously theoretical argument very concrete. If all three keys in your 2-of-3 came from the same manufacturer running the same firmware, a single vendor-wide flaw can compromise the threshold. Sourcing keys from different manufacturers — for example one Coldcard, one Trezor, one BitBox — means a flaw in any one product line leaves you with enough uncompromised keys to stay safe and to migrate.
The trade-offs
Multisig is not free. The costs are real and worth understanding before committing:
- Setup complexity. You need coordinating software (Sparrow, Nunchuk, Caravan and others are commonly used) and you must back up not just the seeds but the wallet descriptor that records the configuration. A seed backup alone is not sufficient to recover a multisig wallet.
- Higher transaction fees. Multisig transactions are larger in bytes and therefore cost more to confirm.
- More things to lose. Three backups in three places means three opportunities for a backup to be misplaced.
- Operational friction. Every spend requires physical access to at least two devices. This is a feature for security and a nuisance for convenience.
- Script exposure on spend. As the Coldcard incident highlighted, the multisig script becomes public the first time you spend from an address. If some of your keys were already compromised, that moment is when the exposure becomes actionable for an attacker.
Is it worth it for you?
A rough framework:
- Under a few thousand dollars: a single hardware wallet with a well-secured seed backup is proportionate. Multisig overhead outweighs the benefit.
- Five figures and up: multisig starts to make clear sense, or at minimum a single-sig wallet protected by a strong BIP-39 passphrase.
- Six figures and up: multisig with vendor diversity and geographic distribution of backups is the reasonable standard.
- Business or shared funds: multisig is effectively mandatory. It is the only way to enforce that no single person can unilaterally move funds.
Getting started safely
If you decide to set up multisig, do it carefully. Create the wallet, back up every seed and the wallet descriptor, then run a full recovery drill on a fresh setup before moving any real funds. Send a small test amount, confirm it arrives, practise spending it back out using the threshold. Only after the whole cycle works end to end should meaningful amounts go in.
The failure mode for multisig is not usually theft — it is users locking themselves out through incomplete backups. Take the drill seriously.
The bottom line
Multisig converts custody from a single point of failure into a system that tolerates failures. Devices break, manufacturers ship bugs, backups get lost, and people make mistakes. A 2-of-3 setup with keys from different vendors absorbs any one of those events without loss.
For anyone holding an amount of Bitcoin that would genuinely hurt to lose, the hours spent setting it up properly are among the best-spent hours in all of self-custody. See also our guides to hardware wallets and seed phrases.