A wrapped token is a token issued on one blockchain that represents an asset locked somewhere else - another blockchain, a custodian, or, in the case of WETH, a contract on the same chain - at a fixed 1:1 rate. Wrapped crypto exists because blockchains and token standards do not talk to each other: Bitcoin cannot be sent to an Ethereum contract, and native ETH cannot be handled by contracts written for ERC-20 tokens. The rule that sorts every wrapped token: what holds the original asset decides the risk. WETH is held by a contract with 2,106,670 ETH in it and 2,058,243 WETH outstanding (Etherscan, September 20, 2026); WBTC’s 116,132 BTC are held by a licensed custodian (wbtc.network, Etherscan, September 20, 2026); a bridged token like WETH.e is held by a bridge.
- Three models, one promise: every wrapped token promises 1:1 redemption; a smart contract (WETH), a custodian (WBTC) or a bridge (WETH.e, USDC.e) is what keeps it.
- WETH is the trust-minimized case: anyone can deposit or withdraw against the WETH9 contract at any time, and the reserve is public: 2,106,670 ETH held against 2,058,243 WETH issued = 102.4% backing on September 20, 2026 (Etherscan).
- Scale of the two big wrappers: WETH has 3,318,447 holders against 193,838 for WBTC - 17 times more addresses, holding 0.62 WETH and 0.60 WBTC per address respectively (Etherscan, September 20, 2026).
- Custodial wrappers depend on people: only DAO-approved merchants such as Galaxy, Amber Group, BitGo, Wintermute, CoinList and Cobo can mint or burn WBTC (wbtc.network).
- Bridged wrappers carry bridge risk: cross-chain bridges lost about $2B across 13 hacks in 2022, 69% of all crypto stolen that year (Chainalysis).
What a Wrapped Token Is
A wrapped token is a claim on an asset that is locked so the token can circulate in its place. The token follows the standard of the chain it lives on - ERC-20 on Ethereum and its L2s, SPL on Solana, BEP20 on BNB Chain - which is what lets it move through contracts that the original asset could not enter. The locked asset stays where it was; the wrapper is a receipt for it, and burning the receipt releases the asset.
Two things are always true of a wrapped token and false of the original: it has a contract address, and it can be approved for spending by another contract. That is the whole reason to wrap. A DEX pool, a lending market or an NFT marketplace needs to call transferFrom() on your balance, and only a token contract can offer that call.
Why Wrapped Tokens Exist
Wrapped tokens exist because two kinds of gaps cannot be crossed directly. The first is between chains: Bitcoin’s ledger has no notion of an Ethereum contract, so BTC that should work in Ethereum DeFi has to be represented by a token that Ethereum contracts understand. The second gap is inside one chain, between a native coin and its token standard: ETH was created in 2015, the ERC-20 standard was finalized in 2017, and ETH does not implement it - a contract that expects transfer(), approve() and balanceOf() cannot receive native ETH (ethereum.org).
The second gap is the one WETH closes. Roughly 3% of the circulating ETH supply sits in the WETH contract for that reason alone (ethereum.org, September 2026): that ETH is not leaving Ethereum, it is being lent to a standard.
How Wrapped Tokens Work: Three Models
Every wrapped token is minted when the original is locked and burned when it is released. What differs is who does the locking.
| Model | Example | Who holds the original | Who can mint and burn | What can break the peg |
|---|---|---|---|---|
| Smart contract | WETH (Ethereum, Base, Arbitrum, Optimism) | The WETH9 contract itself | Anyone, by calling deposit() or withdraw() | A bug in the contract; none found since December 2017 |
| Custodial | WBTC on Ethereum | A licensed custodian (BitGo) | DAO-approved merchants only | Custodian failure, frozen redemptions, regulatory action |
| Bridged | WETH.e on Avalanche, USDC.e on Arbitrum | The bridge contract on the source chain | The bridge, on each crossing | A bridge exploit; the wrapper stays but the reserve is gone |
Sources: Etherscan (WETH9 verified source, creation December 2017), wbtc.network (custody and merchant model), Chainalysis (bridge losses 2022).
Smart-Contract Wrapping: WETH
WETH is minted by the WETH9 contract when ETH is sent to its deposit() function and burned by withdraw(), which returns the same amount of ETH to the caller. No one approves the operation and no one can refuse it: the contract has no owner, no pause switch and no upgrade path, and its source has been verified on Etherscan unchanged since December 2017. The reserve is auditable by anyone at any moment - the ETH balance of 0xC02aaA39b223FE8D0A0e5C4F27eAD9083C756Cc2 is the backing. This is why WETH is the reference case for the whole class: the peg is enforced by code, not by a promise (what WETH is on the home page covers the token itself). How the wrap and unwrap calls work step by step is on the home page, and the unwrap side in detail is in How to unwrap WETH.
Custodial Wrapping: WBTC
WBTC is minted when a merchant delivers BTC to the custodian and burned when the custodian releases BTC to a merchant; end users buy and sell WBTC on the market rather than minting it. On September 20, 2026 the contract shows 116,132 WBTC outstanding across 193,838 holders - 0.60 WBTC per address - with 8 decimals, matching Bitcoin’s (Etherscan). The 1:1 backing is published as an on-chain proof of reserves (wbtc.network), but redemption itself runs through identity-verified institutions approved by DAO governance - the peg holds as long as those institutions and the custodian keep operating. That is a different guarantee from WETH’s: verifiable, but not permissionless.
Bridged Wrapping: WETH.e and Friends
A bridged wrapper is minted on the destination chain when the original is locked in a bridge contract on the source chain. WETH.e on Avalanche is Ethereum WETH locked in the Avalanche Bridge; USDC.e on Arbitrum was USDC locked in the Arbitrum bridge before Circle issued native USDC there. The suffix marks the model: the token lives at its own contract on the new chain and its reserve lives in a bridge. Per Chainalysis, cross-chain bridges lost about $2B across 13 hacks in 2022 alone - 69% of all crypto stolen that year, with Ronin ($625M, March 2022) and Wormhole ($320M, February 2022) the largest - and in every such case the wrapped tokens kept circulating while their backing was gone.
Canonical vs Bridged: Two Wrapped Tokens With the Same Name
The same asset often has two wrappers on one chain, and telling them apart matters more than the name. On Base, WETH at 0x4200000000000000000000000000000000000006 is the canonical wrapper: an OP Stack predeploy that wraps the chain’s own ETH, with 241,555 WETH outstanding and verified WETH9 source (Basescan, docs.base.org, September 20, 2026). On BNB Chain, the token called ETH at 0x2170…33F8 is a bridged peg-token issued by the Binance bridge, not a WETH9 deployment. Both are wrapped ETH; only the first can be unwrapped into ETH on the spot.
- Canonical wrapper: deployed by the chain itself or by the WETH9 code; unwraps into the native asset of that chain.
- Bridged wrapper: deployed by a bridge; “unwrapping” means bridging back, which takes the bridge’s time and fee.
- Lookalike receipt tokens: aWETH (Aave), pETH (JPEG’d) and similar are not wrappers of ETH at all but positions in a protocol; the lookalikes table lists them.
The canonical and bridged WETH contracts per network are on the contract addresses section; each entry says which model it is.
What a Wrapped Token Cannot Do
- It cannot pay gas: network fees are charged in the native coin - ETH, POL, AVAX - never in a wrapped token, including WETH.
- It cannot cross chains by itself: WETH on Ethereum and WETH on Base are separate contracts with separate balances; moving between them is a bridge operation.
- It cannot be more than 1:1: a wrapper never earns yield on its own. Tokens that grow against ETH, such as wstETH or cbETH, are staking receipts, not wrappers.
- It cannot be trusted by name: any contract can call itself “Wrapped Ether”. The contract address and the verified source on the explorer identify the real one; the name does not.