Multisig and multi-share backups: what is the difference?

Both arrangements can use language such as “two of three”, but they put that threshold in different places. With multisig, the threshold governs transaction authorization. With a multi-share backup, it governs recovery. Understanding that distinction matters more than counting devices or pieces of paper.

One threshold controls spending; the other controls recovery

In a typical Bitcoin two-of-three multisig wallet, three independent signing keys are configured and two must authorize a transaction. BIP-383 describes multisig output descriptors, which express signing policies. The keys may be held by one person or distributed among people or services. Three devices loaded with the same key are not three independent signers.

A SLIP-39 multi-share backup instead divides recovery material using a defined secret-sharing scheme. In a simple two-of-three arrangement, two distinct compatible shares can reconstruct the wallet secret. Once recovered into a conventional single-signature wallet, that wallet can sign without fetching the shares again. The shares are not ongoing transaction approvals.

Keep the records that match the design

For multisig, preserve the wallet configuration in the format your coordinator supports. That may include a descriptor, cosigner public keys, derivation information and script type. These records do not normally grant spending power alone, but they can reveal addresses and transaction history. Treat them as private rather than publishing them.

For multi-share recovery, document the standard, threshold, compatible devices and locations through a separate access plan. A twenty-word backup can be a single-share SLIP-39 backup, so word count does not establish a threshold. See wallet backup formats before assuming different cards belong together.

Choose for the problem you actually have

Multisig may suit shared approvals or separation of signing authority, but it adds coordination and recovery records. Multi-share backups may suit distributing recovery risk while retaining a familiar spending process. Neither is universally better, and combining them can make an ordinary holder’s recovery harder to understand.

Start by identifying single points of failure. Consider unavailable people, inaccessible locations, lost passphrases and discontinued software alongside theft. A threshold that is hard for an attacker to meet must still be reachable by you when something goes wrong.

Rehearse the explanation before changing the arrangement. A recovery rehearsal can reveal missing records without gathering all secrets or resetting a working device. If the proposed design cannot be explained clearly to the people who must use it, simplify the plan first.