L2 stablecoin liquidity determines which chains can support large trades, treasury operations, and payments at scale. For the broader L2 landscape, see the Ethereum L2 comparison guide. Pulled September 15, 2026, DefiLlama's stablecoin chain data shows about $14.0 billion in stablecoin float across the eight largest Ethereum scaling chains, with Base ($4.97B) and Arbitrum One ($4.14B) holding roughly 65% of it. The gap between the leaders and the rest is not marginal: Scroll carries $9M and World Chain $20M on the same measure, so an institutional-size ticket has nowhere to clear on those chains.
​

This guide ranks the top L2s by stablecoin liquidity and breaks down the composition per chain. Every figure below comes from DefiLlama's stablecoin chain data, pulled September 15, 2026. Pool-level depth and lending rates move hour to hour, so this guide points you at the live sources for those rather than freezing a number.
​

Stablecoin liquidity is more nuanced than the headline TVL number. Three measurements matter for different use cases:
​

Total stablecoin float on chain. The dollar value of all stablecoins held by addresses on the L2. This is the broadest measure and includes balances in wallets, treasuries, and protocols that may not be actively tradeable. Useful for sizing the chain's "stablecoin economy."
​

DEX-routable liquidity. Stablecoins inside AMM pools that can be swapped on demand. This is what determines slippage on a given trade size. A chain might hold $4B in stablecoins but have only $800M actively in DEX pools.
​

Lending market depth. Stablecoins available for borrow on lending protocols (Aave, Compound, Morpho, Spark). Not directly tradeable but indicates demand for stablecoin yield and capital availability for leveraged operations.
​

This guide covers all three. Most rankings below use total chain float unless otherwise noted, since that's the data DefiLlama publishes consistently.
​

The following ranks the eight largest Ethereum scaling chains by total stablecoin float, per DefiLlama's stablecoin chain data, pulled September 15, 2026. The CCTP column follows Circle's supported blockchains list.

USDC leads overall but not everywhere. Across those eight chains it accounts for roughly 62% of stablecoin float, with USDT at about 17% and DAI at about 4%, per DefiLlama's stablecoin chain data on September 15, 2026. USDT actually outweighs USDC on Mantle and on OP Mainnet, so "USDC is the default" is a Base and Arbitrum fact rather than a universal one.
​

For trade execution, individual pool depth matters more than chain-wide float. Pool depth changes by the hour as liquidity providers enter and exit, so the only honest way to quote it is to read it live: DefiLlama Yields lists per-pool TVL and yield for Curve, Uniswap, Aerodrome, Velodrome and the rest, filterable by chain and by token. Check the specific pool you intend to route through before sizing a ticket, and check it again on the day you trade.
​

The structural point survives any particular snapshot: stable-to-stable depth tracks chain-wide float closely, so the chains at the top of the table above are the chains where a large stablecoin swap clears cheaply, and the chains at the bottom are not viable for size at all no matter which venue you pick.
​

Stablecoin lending markets are the second axis of liquidity that matters. Aave v3 and Morpho carry most L2 stablecoin lending, and borrow rates move with utilization on every block, so a published APY is stale the day it ships. Read the current per-chain USDC markets on DefiLlama Yields before you size a borrow.
​

The shape of the curve is predictable even when the level is not. Smaller chains carry less supplied USDC, so the same borrow demand pushes utilization and therefore rates higher. For applications that borrow stablecoins as part of normal operations, the large-float chains in the table above are consistently the cheaper places to do it.
​

The stablecoin mix on each chain reflects its user base and integrations.
​

USDC dominance on Base. Base holds $4.27B of USDC, about 86% of its $4.97B stablecoin float, with Ethena's USDe a distant second at $367M and USDT not in the top three at all (DefiLlama's stablecoin chain data, September 15, 2026). Coinbase's distribution to Base wallets is USDC-first. For applications building on Base, designing around USDC simplifies the experience but limits cross-stable optionality.
​

Balanced on Arbitrum. Arbitrum's $4.14B splits roughly 56% USDC ($2.32B), 20% USDT ($811M) and 8% PYUSD ($325M), with smaller positions in USDai, USDX and USDS (DefiLlama's stablecoin chain data, September 15, 2026). The diversity reflects Arbitrum's longer history and broader DeFi integrations. Stablecoin venues on Arbitrum see meaningful USDC to USDT flow that Base does not.
​

USDT weight on OP Mainnet. OP Mainnet is one of the few L2s where USDT outweighs USDC: $251M against $180M, roughly 49% against 35% of a $513M float (DefiLlama's stablecoin chain data, September 15, 2026). DAI, often assumed to be large here, is $14M, under 3%.
​

Thin floats on the ZK rollups. Linea ($37M), ZKsync Era ($38M) and Scroll ($9M) each hold less stablecoin value than a single mid-size lending market on Arbitrum (DefiLlama's stablecoin chain data, September 15, 2026). Composition on chains this small is dominated by whichever protocol launched most recently, so it tells you little about user preference.
​

Circle's Cross-Chain Transfer Protocol changes which chains effectively share USDC liquidity. Pre-CCTP, USDC on each chain was issued as a wrapped representation by Circle plus the bridge, bridged USDC and native USDC were not fungible.
​

Post-CCTP, USDC on supported chains is native: any USDC on Arbitrum is fungible with any USDC on Base, with chain transfer happening via burn-and-mint at no slippage. Circle's supported blockchains list now runs to roughly thirty chains, including Arbitrum, Base, OP Mainnet, Polygon PoS, Linea, Ink, Starknet, Unichain, World Chain and Solana. Mantle, Scroll and ZKsync Era are not on it.
​

For applications that operate in USDC across multiple L2s, this effectively creates a unified $30B+ liquidity pool. The treasury or trading desk doesn't need to maintain pre-positioned USDC on every chain, it can rebalance via CCTP at near-zero cost. The chains without CCTP, Scroll and ZKsync Era among them, sit outside this unified pool.
​

Pool depth tells you what is possible; daily volume tells you what is actually flowing. DefiLlama publishes DEX volume per chain and per venue and updates it daily, which is the right place to read it rather than a figure frozen into a guide.
​

Volume correlates with float but adds the dimension of active turnover. A chain with deep pools but thin volume can absorb size yet sees less price discovery, so quotes drift further from fair value between trades. Higher-volume chains tend to hold tighter spreads even at smaller trade sizes.
​

Beyond plain stablecoins, yield-bearing variants have taken real share on L2s. Ethena's USDe is the clearest case: $367M on Base and $60M on Mantle, against $4M on Arbitrum, per DefiLlama's stablecoin chain data on September 15, 2026. Sky's USDS sits at $100M on Arbitrum and almost nothing on Base on the same pull, so the two largest yield-bearing stables have close to opposite chain footprints.
​

Yields on the staked wrappers (sUSDe, sUSDS) float with funding rates and the Sky Savings Rate respectively and are published live on DefiLlama Yields. Treat any quoted APY older than a week as decorative.
​

For applications that need to hold or accept stablecoin reserves, the choice between non-yield (USDC) and yield-bearing (sUSDe, sUSDS) variants is increasingly architectural. Non-yield is simpler operationally; yield-bearing earns a meaningful return on idle balances.
​

The liquidity ranking matters for three categories of decision.
​

Choosing a primary L2 for stablecoin-heavy apps (the ZK rollup comparison covers chains with faster L1 finality). If stablecoin depth is critical (payments, lending, perps with stablecoin collateral), Base, Arbitrum and Polygon PoS are the only Ethereum scaling chains above $3B of float on the September 15, 2026 pull from DefiLlama's stablecoin chain data. ZK rollups can serve specific use cases but won't carry institutional-size flows yet.
​

Designing multi-chain architectures. Treasury or trading apps that need to operate across multiple chains should expect uneven liquidity. A common pattern: hold the bulk of stablecoin reserves on Arbitrum or Base, route operations through CCTP to other chains as needed, accept higher slippage on smaller-chain swaps.
​

Pricing trades and slippage estimates. Pre-trade slippage estimates should pull from per-chain pool depth, not chain-wide TVL. A $500K stablecoin swap on Linea will hit different slippage than the same swap on Arbitrum, even if both chains show similar headline TVL.
​

Eco is a stablecoin execution network that orchestrates stablecoin flows across the 15 chains it supports, including Arbitrum, Base, OP Mainnet, zkSync Era, Linea, Scroll, Polygon, and Solana. Eco Routes (CLI + API) accepts a transfer or swap intent and routes through whichever combination of CCTP, third-party bridges, DEX aggregators, and OTC liquidity sources clears at the best price. For applications operating across L2s with uneven stablecoin liquidity, Routes abstracts the chain selection: the integration sees a single API; the network handles "swap on Arbitrum where liquidity is deep, settle to Linea via CCTP" automatically.
​

Base leads with $4.97B of stablecoin float, ahead of Arbitrum One at $4.14B, per DefiLlama's stablecoin chain data pulled September 15, 2026. Together they hold roughly 65% of the float across the eight largest Ethereum scaling chains. Single-pool depth moves too fast to quote here: read it live on DefiLlama Yields.
​

No. Circle's supported blockchains list covers roughly thirty chains, including Arbitrum, Base, OP Mainnet, Polygon PoS, Linea, Ink, Starknet, Unichain and World Chain. Scroll, Mantle and ZKsync Era are not on it and use bridged USDC. Native USDC is fungible with USDC on any other CCTP chain via burn-and-mint.
​

There is no fixed answer, and any guide that gives you one is quoting a pool that has since changed. The rule that holds is proportional: slippage stays negligible while your trade is small against pool depth and steepens sharply above roughly 5% of it. So the workable method is to look up the depth of the specific pool on DefiLlama Yields, or to pull a live quote from an aggregator, and size against that number rather than against a chain-level float figure.
​

Coinbase's user distribution to Base is USDC-by-default. Coinbase's wallets, exchange withdrawals, and consumer apps push users toward USDC. USDT exists on Base but lacks an equivalent distribution channel. USDC holding roughly 86% of Base's stablecoin float, with USDT outside the top three, reflects Coinbase's role as Base's primary user-acquisition channel (DefiLlama's stablecoin chain data, September 15, 2026).
​

Whichever chain has the most USDC supplied relative to borrow demand at the moment you ask, which in practice has meant Base or Arbitrum. Rates reprice every block, so check the live per-chain USDC markets on DefiLlama Yields rather than trusting any published figure, including this one.
​

Standard Transfers are free and Fast Transfers cost 0 to 13 basis points depending on the source chain, per Circle's CCTP fees page. On top of whatever Circle charges, the user pays gas on the source chain to initiate the burn and gas on the destination chain to claim the mint, and waits for the attestation, whose duration depends on the source chain's finality.
​

Yield-bearing stables introduce smart-contract and protocol risk on top of issuer risk. sUSDS inherits Sky's Maker-derived risk profile; sUSDe carries Ethena's funding-rate and basis risk. Holdings on L2 add the rollup's security stage. For most treasury operations, holding under 25% of stable reserves in yield-bearing variants is a common risk budget, the rest stays in plain USDC or USDT.
​

Arbitrum had a 2-year head start over Base and accumulated DeFi protocols across two market cycles. Base benefits from Coinbase's distribution channel pushing USDC into Base wallets. Other L2s lack equivalent distribution or DeFi mass. The concentration is a stable equilibrium for now, most institutional flows stay where the liquidity already sits.
​

USDT0 is an omnichain representation of Tether built on LayerZero's OFT standard, and DefiLlama tracks it as its own asset alongside USDT. On the L2s covered here, USDT and its omnichain form together are the second-largest stablecoin block after USDC, roughly 17% of float across the eight chains (DefiLlama's stablecoin chain data, September 15, 2026), and the clear leader on Mantle and OP Mainnet. Unlike CCTP, USDT0 does not burn and mint through an issuer attestation; LayerZero's messaging layer coordinates supply across chains.
​

Less than the top of the table suggests, but more than most guides admit: Base and Arbitrum have traded the lead between them while staying far ahead of everything else, and the two sat within $1B of each other on the September 15, 2026 pull from DefiLlama's stablecoin chain data. Bigger movements happen at the edges as newer chains capture flow from emerging use cases. Re-checking quarterly is reasonable for most builders; weekly checks are useful only if you trade enough size that pool depth changes affect execution.
​