Liquid staking has become the foundation of DeFi yield. With over 31% of all ETH staked and ~40% of that flowing through liquid staking tokens (LSTs), the choice of which LST to hold has portfolio-level consequences. Following the Pectra hard fork and the rise of restaking via EigenLayer, the LST landscape now includes both pure staking products and yield-stacked liquid restaking tokens (LRTs). Here are the seven leading ETH liquid staking protocols ranked by total value locked, security model, and effective yield, with notes on validator decentralization and integration depth.

1. Lido stETH (LDO)

Lido remains the largest LST by TVL with ~28% of all staked ETH. stETH is integrated across virtually every Ethereum DeFi protocol and serves as the de facto interest-bearing collateral asset.

  • Why it matters: Deepest secondary liquidity (Curve stETH/ETH is one of the largest DEX pools on Ethereum); broad protocol integration.
  • Key risk: Validator set is permissioned (curated set of node operators); LDO governance is concentrated; staking ratio dominance creates a perceived centralization concern for Ethereum itself.
  • Coverage: live profile · prediction

2. Rocket Pool rETH (RPL)

Rocket Pool is the most decentralized LST: anyone can run a minipool with 8 ETH plus RPL collateral. rETH is the value-accruing token (price increases over time rather than rebasing).

  • Why it matters: Permissionless validator participation; smallest centralization concern of any major LST; conservative engineering culture.
  • Key risk: Higher operational complexity for solo node operators; smaller TVL means thinner secondary liquidity; rETH price premium over ETH varies.
  • Coverage: live profile · prediction

3. Frax sfrxETH (FRAX)

Frax’s LST splits the staking position: frxETH (1:1 with ETH, no yield) and sfrxETH (concentrates the staking yield from validator operations).

  • Why it matters: Higher headline APR than peers because non-stakers (frxETH holders) cede their yield to sfrxETH; tight integration with the Frax ecosystem (Curve, Convex).
  • Key risk: Reliance on Frax governance; validator operations are run by Frax-affiliated entities; less liquid than stETH on the open market.

4. StakeWise V3 osETH (SWISE)

StakeWise V3 is a permissionless vault architecture: each operator runs their own vault with custom parameters, and osETH is the aggregate-backed liquid token.

  • Why it matters: Most customizable staking architecture; suitable for institutional operators who want isolated vault risk.
  • Key risk: Smaller market share means lower DeFi integration; smart-contract complexity per vault.

5. ether.fi eETH (ETHFI)

Ether.fi is the largest non-custodial LST and the first major LST to natively restake on EigenLayer. weETH (wrapped eETH) is integrated across major DeFi protocols.

  • Why it matters: Native restaking exposure (EigenLayer points + AVS yield + base ETH staking yield); the largest LRT by TVL.
  • Key risk: EigenLayer restaking adds slashing and AVS counterparty risk; ETHFI tokenomics are still maturing.
  • Coverage: live profile · prediction

6. Mantle LSP mETH (MNT)

Mantle LSP’s mETH is positioned as a yield-stacked LST: native Ethereum staking + Mantle ecosystem incentives. Backed by the Mantle treasury.

  • Why it matters: Treasury-subsidized yield is the highest in the LST category; tight integration with Mantle L2 DeFi.
  • Key risk: Yield is partially subsidized — not all of it is sustainable from validator operations; concentrated governance via the BitDAO/Mantle treasury.
  • Coverage: live profile · prediction

7. Swell swETH (SWELL)

Swell focuses on the LRT and pre-confirmation use case. swETH is a vanilla LST; rswETH (restaked variant) targets EigenLayer-aligned yield.

  • Why it matters: Builder-friendly: Swell’s L2 (built on Polygon CDK) is positioned as a restaking-native execution layer.
  • Key risk: Smaller TVL; ecosystem traction depends on the success of the Swell L2 roadmap.

The restaking dimension: where LRTs fit

Restaking via EigenLayer has been the largest single development in the ETH staking landscape since the Merge itself. Liquid Restaking Tokens (LRTs) — eETH, rsETH, ezETH, pufETH, swETH, and others — let users earn the base ETH staking yield, EigenLayer points or AVS yield, and (typically) restaking-protocol token rewards stacked on top.

The trade-off: every additional AVS opted into adds slashing risk. The slashing parameters for AVSs are still being defined and have not been fully battle-tested at scale. A meaningful slashing event could permanently reduce the value of an LRT below 1:1 ETH. For users, the right rule is to understand exactly which AVSs the LRT operator has opted into and what their respective slashing conditions are.

For institutional allocators, the conservative path remains pure LSTs (stETH, rETH, sfrxETH) where slashing risk is bounded to the underlying validator set behavior. LRTs are appropriate only after explicit acceptance of the additional risk surface — and only for a portion of the total ETH allocation.

Methodology

Yields shown are 30-day moving averages net of protocol fees, pulled from each protocol’s public dashboard. TVL is sourced from DefiLlama and validated against on-chain contract balances. Security considerations include validator set composition (permissioned vs permissionless), slashing-loss insurance, and the maturity of the redemption queue. LRTs (eETH, rswETH) carry restaking risk in addition to base validator risk and are not directly comparable to pure LSTs.

Concentration risk and the case for diversification

Lido controls roughly 28% of all staked ETH. That number has fluctuated over the years but remains the largest single concentration in Ethereum’s consensus layer. Ethereum researchers have repeatedly warned about the centralization risk: if any single staking entity exceeds 33% of stake, that entity gains the ability to delay or halt finalization. Above 50%, it can censor transactions. Above 66%, it can finalize alternative histories.

For users, the practical implication is that diversifying staked ETH across multiple LSTs (and ideally including a direct solo-staking position if technically feasible) is a public good as well as a personal risk-management practice. Rocket Pool and ether.fi explicitly position themselves as decentralization-aligned alternatives to Lido. The trade-off is typically modestly lower yield and thinner secondary liquidity.

The “Should I stake at all?” question is genuinely worth asking. Locked ETH cannot be sold quickly during volatility spikes. Withdrawal queues during periods of stress have historically extended to multiple days. The yield from staking (~3-4% net) is meaningful but not transformative; treat it as a base layer, not a primary investment thesis.

Frequently asked questions about ETH liquid staking

Is liquid staking better than solo staking?

For most users, yes — solo staking requires 32 ETH, technical operational competence, and continuous uptime. Liquid staking accepts any amount, requires no technical setup, and returns a fungible token usable across DeFi. The trade-off is the smart-contract risk and centralization concerns of the LST protocol. Solo staking remains the gold standard for security and decentralization for technically capable users with sufficient ETH.

What is the difference between stETH and rETH?

stETH (Lido) is a rebasing token — your balance increases daily as staking rewards accrue. rETH (Rocket Pool) is a non-rebasing value-accruing token — the balance stays the same but the redemption value of each rETH increases over time relative to ETH. Both designs are functionally equivalent; rETH is generally simpler for tax accounting and for DeFi protocols that don’t handle rebasing well.

What happens during withdrawal queues?

Ethereum has a built-in validator exit queue that throttles how quickly stake can be unwound. During the 2023 Shapella upgrade and again during periods of price volatility, the exit queue has extended to multiple days. LST protocols handle this with internal liquidity buffers — you can usually swap LST for ETH instantly on a DEX, but at a small discount during stress periods. Direct LST redemption back to ETH via the protocol can take hours to days.

Are LRTs worth the extra risk?

Depends on your risk tolerance and time horizon. LRTs offer higher headline yield by stacking EigenLayer AVS rewards on top of base staking yield, but they expose you to additional slashing surface area. For 2026, the consensus advice is that LRTs are appropriate for sophisticated users with a deliberate restaking thesis, but not as a default substitute for pure LSTs. Diversification across LSTs (e.g., 50% stETH + 30% rETH + 20% LRT exposure) is a reasonable middle ground.

Bottom line: which LST should I hold?

For most users with a meaningful ETH allocation, the practical answer in 2026 is: a 60/30/10 split across stETH, rETH, and a chosen LRT. stETH provides liquidity and broadest DeFi integration. rETH provides the strongest decentralization story and conservative engineering. A small LRT allocation captures restaking upside without concentrated slashing exposure.

For users who prioritize maximum decentralization, the inverse split (60% rETH, 30% solo staking if feasible, 10% stETH) makes more sense. For users who prioritize yield over all else, sfrxETH or ether.fi eETH will likely produce the highest headline APR but with more concentrated counterparty exposure.

Whatever you choose, monitor: (1) Validator queue depth and exit queue depth on beaconcha.in. (2) The DEX-vs-protocol redemption discount for your LST during stress periods. (3) Any slashing events involving your LST’s validator set. (4) Governance proposals affecting protocol parameters.

Liquid staking is one of the highest-quality yield products in crypto — sustainable, transparent, and tied to real Ethereum economic activity. Used appropriately, it should form a meaningful part of any long-term ETH allocation. Used carelessly (chasing the highest APY without understanding the risk surface), it can amplify losses during stress periods.

For more rankings and analysis, see all analysis →