When you open a self-custody wallet for the first time, a lot of new words arrive at once: private key, public key, address, recovery phrase. They sound like four separate things to keep track of. In reality they are one chain of ideas, each derived from the one before it. Once you see how they connect, self-custody stops feeling like a mystery and starts feeling like a lock you understand.
This article explains public and private keys in plain language, with no cryptography background required.
Two keys that belong together
A self-custody wallet is built on a pair of keys that are mathematically linked.
The private key is a very large secret number. It is the real wallet. Whoever knows the private key controls the funds, and nobody who does not know it can move them. This is the one thing in the entire system you must never share, never type into a website, and never store where someone else could read it.
The public key is calculated from the private key using a one-way piece of math. One-way means it is easy to go from the private key to the public key, but effectively impossible to go backward from the public key to the private key. Because of that, the public key is safe to reveal. Anyone can hold it without gaining any power over your money.
The relationship only runs in one direction. Your private key produces your public key, but your public key can never expose your private key. That asymmetry is the whole foundation of self-custody.
Where your address comes from
You rarely see the raw public key. What you share with other people is your address, the string that starts with 0x on Base and other Ethereum networks. The address is a shortened, tidied version of your public key, produced by hashing it and keeping part of the result.
So the chain of derivation looks like this:
- Private key, the secret number
- Public key, calculated from the private key
- Address, calculated from the public key
Each step is one-way. Someone who knows your address cannot work out your public key in a way that helps them, and certainly cannot reach your private key. That is why it is completely safe to post your address, put it in your profile, or send it to a friend so they can pay you. An address is like an account number you can hand out freely. The private key is the signature that authorizes spending, and that stays with you alone.
What signing actually means
Here is the part that makes the design useful. When you approve a transaction in your wallet, you are not sending your private key anywhere. Instead, your wallet uses the private key to produce a signature, a unique stamp that only that specific key could have created for that specific transaction.
The network can then check the signature against your public key and confirm two things: that the transaction was authorized by the holder of the private key, and that not a single detail was altered afterward. It confirms all of this without ever seeing the private key itself.
Think of it as a wax seal that anyone can verify but only you can press. This is why you can safely broadcast a transaction across a public network. The proof of ownership travels with it, but the secret that created the proof never leaves your device.
Where the recovery phrase fits
If a private key is a long secret number, remembering it or typing it by hand would be painful and error prone. The recovery phrase, usually twelve or twenty four words, solves that. It is a human readable form of a master secret that your wallet turns into private keys.
Two useful consequences follow from this.
First, one recovery phrase can generate many keys and many addresses. That is how a single wallet app can hold several accounts. They all descend from the same phrase.
Second, the recovery phrase deserves exactly the same protection as a private key, because it can regenerate your private keys on any device. Anyone who reads your phrase can rebuild your wallet and take everything. There is a longer companion article on storing that phrase safely, and it is worth reading.
Why this matters for holding your own money
In a bank, the bank holds the secret that controls your account, and you ask it to act for you. In self-custody, you hold the secret. That is the trade at the center of crypto.
The upside is real. No company can freeze your wallet, no account can be closed on you, and no third party can lose your funds through their own failure. You transact directly, and the network verifies you by your signature rather than by a login.
The responsibility is equally real. There is no support desk that can reset a lost private key or recovery phrase, because the whole point is that only you ever had it. If the secret is lost, the funds are unreachable. If the secret is stolen, the funds are gone. No one can reverse either outcome.
That is not a flaw to be fixed. It is the direct result of the two-key design that lets you hold money nobody else can touch.
The short version
- Your private key is the wallet. Keep it secret, always.
- Your public key is derived from it and is safe to reveal.
- Your address is derived from the public key and is safe to share.
- Signing proves you authorized a transaction without exposing the private key.
- Your recovery phrase can rebuild your private keys, so guard it like the key itself.
Understand these five points and you understand how self-custody works. Everything else your wallet does is built on top of them.