How to Stake ETH on Base and Arbitrum: Lido, Rocket Pool, and What the Gas Actually Costs

What staking on an L2 actually means

First, the thing that confuses everyone: you cannot run an Ethereum validator from Base or Arbitrum. Validators live on Ethereum mainnet, lock 32 ETH each, and sign blocks — that role is anchored to L1 and cannot move to a rollup. What "staking on L2" means is holding liquid staked ETH tokens — stETH, rETH, wstETH, cbETH — that represent a claim on staked ETH running on mainnet. The validators stay on L1; you hold the receipt on L2.

Two protocols dominate the supply of those receipts: Lido (the largest, issuing stETH and its wrapped version wstETH) and Rocket Pool (issuing rETH). Both are deployed on Base and Arbitrum, both let you swap native ETH for the liquid staked token in a single transaction, and both let you redeem the token back to ETH when you want out. Coinbase also issues cbETH, which works the same way and is especially common on Base because of the Coinbase integration.

If you are new to liquid staking generally, the broader DeFi landscape on L2 is covered in the best L2 for DeFi comparison. This article focuses on staking specifically.

Why L2 staking is materially cheaper than mainnet

On Ethereum mainnet, swapping ETH for stETH through Lido costs about $5-15 in gas on a quiet day, $30+ on a busy one. The same swap on Base costs about $0.01-0.02 — roughly 500 to 1,000 times cheaper. Redeeming stETH back to ETH shows the same gap. Over a year of active position management, the difference compounds: a staker who adjusts, redeems, re-stakes, or moves positions a few times a month can pay hundreds of dollars in mainnet gas versus cents on Base.

The yield itself — currently in the 2.5-4% APY range depending on network conditions — is the same whether you hold stETH on mainnet or wstETH on Base. Lido and Rocket Pool route validator rewards to the same pool regardless of which chain the receipt token lives on; the chain only affects the transaction cost of acquiring, holding, and redeeming the receipt. So if your goal is to earn yield on ETH you already hold, the L2 route is strictly better on costs.

The arithmetic on the gas reserve you need for a year of active staking management is in the gas reserve guide — a $10 ETH buffer on Base covers hundreds of staking transactions.

Lido vs Rocket Pool vs cbETH on Base

Three liquid staked ETH tokens dominate on Base, and they are interoperable but not identical:

All three trade at roughly the same effective yield because they all represent the same underlying validator rewards. The choice is usually about which one your preferred DeFi protocol accepts as collateral, which has the deepest liquidity on the DEX you use, and which issuer you trust. Lido is the largest by market share, which is a feature (deepest liquidity, most integrations) and a concern (single largest staking pool). Rocket Pool has the most decentralized operator set. cbETH is the most Coinbase-integrated.

A practical note: wstETH is the wrapped version of stETH that is actually used on L2s. stETH on mainnet rebalances daily to reflect accrued rewards, which causes accounting headaches on other chains; wstETH instead appreciates in value, which is simpler. If you see "stETH" referenced in mainnet context and "wstETH" in Base context, they are the same underlying asset.

The actual gas cost of each staking action

Receipt-level costs on Base under normal 2026 conditions, based on the live tracker and my own transactions:

The same actions on Arbitrum cost within a cent of these — both chains sit in the same sub-cent-to-cents fee regime. The fee comparison details are in the Base vs Arbitrum gas guide. The short version: for staking specifically, it does not matter which L2 you pick; pick the one you already use.

Using staked ETH as DeFi collateral

The reason liquid staked ETH became a foundational DeFi primitive is that it lets you earn staking yield and use the same ETH as collateral for borrowing. On Base, Aave accepts wstETH as collateral — you can supply wstETH, borrow USDC against it, and use the USDC in a yield strategy while your original ETH keeps earning validator rewards. The full Aave workflow is in the Aave on Base guide, and the lending-side fee breakdown is in the Aave fees guide.

This compounds into a strategy sometimes called "loop staking" or "leverage staking": supply wstETH, borrow ETH against it, swap the borrowed ETH for more wstETH, supply that, repeat. Each loop costs about $0.05-0.10 in Base gas (approve, supply, borrow, swap, supply). On mainnet the same loop costs $30-60 per iteration, which makes small positions uneconomical — on Base you can meaningfully loop a $500 stake. The risk is the same as any leverage: if ETH drops sharply, your position can be liquidated. Keep loan-to-value under 50% and you are safe in almost any market conditions.

One ERC-20 approval note before the first DeFi interaction: every protocol you supply wstETH to needs a one-time approval transaction to access your tokens. The approval guide explains why this step exists and why you should revoke approvals you no longer use.

The risks that are not in the Lido dashboard

Liquid staking on L2 has the same underlying risks as on mainnet, plus a layer of L2-specific risk:

• Slashing risk. If the validators backing stETH or rETH are slashed (penalized for misbehavior or downtime), the value of the receipt token drops. Lido and Rocket Pool have multi-sig operator sets to mitigate this, but the risk is never zero.
• Withdrawal queue delays. Redemption back to ETH goes through a queue. Under heavy redemption pressure (typically during market panics), the queue can back up for days or weeks. Your funds are not stuck — but they are not instant.
• Smart contract risk. Lido and Rocket Pool contracts are heavily audited, but any vulnerability would affect the receipt token across all chains, not just mainnet.
• L2 bridge risk for the receipt token. wstETH on Base is bridged from mainnet. If the canonical bridge between Base and Ethereum failed (an extreme scenario), the L2 representation could become disconnected from the L1 asset. This is the same risk every bridged asset carries; the canonical Superchain bridge minimizes it but does not eliminate it.
• Liquidity risk. If you need to exit wstETH quickly without using the redemption queue, you swap it on a DEX. Under market stress, slippage can be significant — you might lose 1-3% versus redeeming directly with Lido.

None of these risks are unique to staking on L2 — the first three apply identically to mainnet staking. The L2-specific risks (bridge and liquidity) are smaller in practice than they sound, because the canonical Superchain bridge is well-tested and Base's wstETH liquidity pool is deep. But the honest framing is: liquid staking is not the same as holding native ETH, and you should understand the failure modes before locking significant value in receipt tokens.

When to stake on L2, and when not to

Stake on L2 when you already hold ETH on Base or Arbitrum and want to earn yield without exiting to mainnet. The cost of acquiring wstETH is cents, the yield is the same as mainnet, and you keep your funds in the ecosystem where you can use them. This is the dominant use case for most retail stakers in 2026 — the gap between mainnet and L2 gas is so large that mainnet staking only makes sense for very large positions where the L1 fee is a rounding error.

Do not stake on L2 if your goal is to run a validator yourself. That role is anchored to L1 — you need 32 ETH on mainnet, dedicated hardware or a solo staking service, and the technical setup has nothing to do with which chain your receipt token lives on. Liquid staking is for earning yield on ETH without running infrastructure; solo staking is for participating in consensus directly. They are different products.

Do not stake on L2 if you might need the underlying ETH on mainnet within a week. The redemption queue plus the L2-to-L1 withdrawal window can take longer than that in stressed conditions. For funds you might need on short notice, hold native ETH.

For the broader question of which L2 to stake on — Base vs Arbitrum vs others — the cheapest L2 ranking and the Base vs Arbitrum comparison have the decision framework. For staking specifically, the answer is almost always "the chain you already use" — the gas difference between them is too small to matter next to the convenience.