What is an address?

An address is what you give someone so they can pay you. It is a short encoding of a lock: the conditions a later spend must meet. Usually that condition is a signature from a key in your wallet. Sharing an address does not share the key.

The formats you will see

Legacy addresses start with 1. Script addresses, used by some multisig and older wallets, start with 3. Native SegWit addresses start with bc1q. Taproot addresses start with bc1p. All four are Bitcoin addresses. Most current wallets can pay all of them. The bc1 formats usually cost less to spend, because their signature data counts less towards the block limit.

A Lightning invoice, which starts with lnbc, is not an on-chain address. Addresses for other coins and networks can look similar. When an exchange asks which network to use for a withdrawal, choose Bitcoin, and never paste a Bitcoin address into a form for a different asset.

Check the whole address

Malware can swap a copied address for one that belongs to an attacker. Attackers can also generate look-alikes that share the first and last few characters with yours. So compare the whole address, middle included. With a hardware wallet, compare it on the device’s own screen. The checksum in an address catches most typing mistakes. It cannot catch a valid address that belongs to someone else.

A fresh address each time

Wallets offer a new address for each receive. They all come from the same seed, so the same backup covers them. Reusing one address works, but it links every payment to it on the public chain.

A wallet can work out a long run of addresses from an extended public key, or xpub, without the private keys. That is how watch-only apps and multisig coordinators show your addresses. It is also why an xpub should be treated as private: it cannot spend, but it reveals every address it makes, and so your whole history.