How to Stake Polkadot (DOT)

How to stake DOT in 2026: native delegation vs liquid staking vs exchange staking, validator selection, slashing risk, and step-by-step delegation. Beginner-friendly.

What you'll need (prerequisites)

• DOT held in a self-custodial wallet
• Wallet that supports polkadot staking
• A small native-token reserve for transaction fees
• A shortlist of validators (use the chain's official dashboard)

Step-by-step

1. Step 1: Acquire DOT

   Buy DOT on a major exchange (Coinbase, Kraken, Binance) or via a DEX. Withdraw to a self-custodial wallet that supports Polkadot.

2. Step 2: Choose a staking method

   Native staking: delegate directly from your wallet to a validator. Liquid staking: deposit into a liquid-staking protocol and receive a yield-bearing receipt token. Exchange staking: easiest UX but custodial. Pick based on your custody and yield preferences.

3. Step 3: Pick a validator / operator on Polkadot

   For native staking, research validator uptime, commission rate (5-10% is typical), self-stake, and slashing history. Avoid concentrating in the largest validators — distribute to support decentralisation.

4. Step 4: Delegate

   In your wallet's staking tab (or the chain's official dashboard), select the validator and enter the amount. Confirm the transaction. There is usually a small network fee.

5. Step 5: Confirm and monitor rewards

   Rewards typically accrue per epoch (5 mins to 24 hrs depending on chain). Check your validator's performance weekly — if uptime drops, redelegate.

6. Step 6: Unstake when needed

   Since Referendum 1910 (enacted 6 July 2026) the nominator unbonding period is roughly 2 days, down from 28. Validators keep the longer period, and unbonds started before the upgrade stay on the old 28-day schedule.

How DOT staking actually works: NPoS, eras and era points

Polkadot uses Nominated Proof-of-Stake (NPoS). Validators run the nodes; nominators — that is you — back a capped list of candidates with bonded DOT, and an on-chain election algorithm allocates nominator stake across the elected set so that backing is spread as evenly as possible. That allocation step is why Polkadot behaves differently from Cosmos-style chains: you do not choose one validator and receive that validator's output, you submit a set of acceptable candidates and the protocol decides which of them your stake ends up behind for each era.

Time is measured in eras of roughly 24 hours (six sessions each). Rewards are computed per era and are not proportional to stake alone: validators accumulate era points for useful work such as authoring blocks and producing valid consensus messages, and the era's reward pool is split by era points first, then shared among the nominators backing each validator in proportion to the stake actually assigned to them, after the validator's commission. Two consequences matter. First, a validator with an enormous stake is not more profitable per DOT — payouts per validator are broadly similar, so the stake bonded behind a heavily-nominated validator earns a thinner slice. Smaller elected validators frequently pay better. Second, rewards must be claimed and are only retained on-chain for a limited window (84 eras), so an unclaimed payout can expire.

Unbonding: the July 2026 change you need to know about

Polkadot's 28-day unbonding period was a defining feature of DOT staking for years, and a great deal of third-party content still describes it as current. It changed. Referendum 1910, enacted on 6 July 2026, reduced the nominator unbonding period from 28 days to approximately two days and removed nominators from slashing exposure entirely. Validators are unaffected: they keep the existing unbonding and slash-deferral periods, remain slashable, and remain subject to the minimum self-stake requirement that a companion referendum raised. The reasoning was straightforward — once validators must post meaningful self-stake and carry direct slashing risk, exposing nominators to the same penalty adds user pain without adding much security.

One transitional detail catches people out: unbonding requests submitted before the upgrade stay on the original 28-day schedule. If you started a long unbond in June 2026 and want the shorter window, you have to rebond and start the process again. Note also that user interfaces, wiki pages and exchange help articles update slowly — if a dashboard still shows 28 days, verify against current chain state before assuming your funds are stuck.

Slashing on Polkadot after Referendum 1910

Historically, nominators shared proportionally in any slash applied to a validator they backed, with the severity scaled by how many validators misbehaved simultaneously — an isolated equivocation was penalised lightly, while a correlated failure across a large share of the validator set was penalised severely, on the theory that correlated failures are what actually threaten the chain. Since Ref 1910, nominators are no longer slashable. Validators still are, for equivocation (signing conflicting blocks) and other consensus offences, and slashes are still applied with a deferral period so governance can intervene in the case of an obvious infrastructure accident.

This does not make validator choice irrelevant. A validator that is slashed, chilled or otherwise removed from the active set stops producing era points, and stake assigned to it stops earning. A validator that raises commission to 100% keeps everything. Nomination pool members should also read pool documentation carefully, since pool accounting historically propagated slashes to members and pool interfaces may lag the runtime change.

Minimums: nominate directly, or join a nomination pool

Direct nomination has a real capital floor. Polkadot documentation puts the minimum to nominate at 250 DOT, and separately warns about the minimum active nomination — a dynamic threshold, driven by how much DOT is staked overall, below which your bond may be elected out and earn nothing at all. Bonding exactly the minimum is therefore a bad idea: the threshold moves, and a nomination that falls under it is inactive until you top it up.

Nomination pools exist precisely to solve this. A pool aggregates many small bonds into one nominating entity, and you can participate with as little as 1 DOT and still receive rewards. The trade-off is that the pool's operator chooses the validator set, you inherit the pool's configuration, and pool membership is a slightly different accounting object than a direct bond (partial unbonds are capped in number, and pool interfaces have their own claim flow). For most retail-sized positions, pools are the correct answer; direct nomination makes sense once your bond comfortably exceeds the minimum active nomination and you want control over which operators you back.

Choosing validators

Because the election spreads your stake, you should submit a set of candidates rather than one. Filter on: era points earned relative to peers (the direct measure of productive work), commission and its change history, self-stake, whether the operator runs a single machine or a diversified professional setup, and total backing — remembering that heavily-backed validators dilute per-DOT rewards. Avoid validators with identity fields left blank and no public operator; avoid nominating multiple validators run by the same operator, which defeats the diversification the election is trying to give you. The staking dashboard's built-in ranking is a reasonable starting point but optimises for expected return, not decentralisation.

Where the yield comes from

DOT staking rewards come from protocol issuance, split between stakers and the on-chain treasury, and distributed by era points as described above. Polkadot's issuance model has been the subject of active governance, and the network has moved toward lower inflation over time — which lowers nominal staking APR while also lowering dilution for holders. The number that matters is not the headline APR but the APR net of commission, net of issuance-driven dilution, and net of the eras your stake spent inactive. Read the current figure from the staking dashboard at the moment you bond, and recheck it after governance changes rather than trusting cached third-party numbers.

Liquid staking alternatives for DOT

Bifrost is the established liquid staking venue in the Polkadot ecosystem; its vDOT is a yield-bearing token that appreciates against DOT as staking rewards accrue and can be traded or used as DeFi collateral rather than sitting locked. The historical case for vDOT was overwhelmingly about escaping the 28-day queue — a case that Ref 1910 substantially weakens for ordinary nominators. What remains is composability: if you want staked DOT that can also be collateral, an LST still does something native staking cannot.

Weigh that against the added layers: smart-contract and cross-chain (XCM) risk, a protocol fee on rewards, delegation decisions made on your behalf, and secondary-market discount risk when you want out quickly. With a two-day native unbond available, the liquidity premium a liquid staking token can justify is smaller than it used to be.

Risks worth pricing in

• Inactive-bond risk. Falling below the minimum active nomination means zero rewards while your DOT is still bonded.
• Unclaimed payouts expire. Rewards are retained for a limited number of eras; claim or use a pool that claims for you.
• Stale information risk. Much of the public web still states a 28-day unbond and slashable nominators. Verify parameters on-chain.
• Commission changes and validator chilling can silently take your effective yield to zero.
• Pool operator risk if you use nomination pools — you inherit their validator selections.
• LST risk if you use vDOT: contract, bridge and market-discount exposure on top of staking.

Sources

• Polkadot Forum — Ref 1910 enacted 6 July 2026: nominators no longer slashable, ~2-day unbonding
• Polkadot Wiki — Staking (NPoS, eras, era points, reward claim window)
• Polkadot Wiki — Nomination pools (1 DOT to join; 250 DOT to nominate directly)
• Polkadot Cloud — Unbonding: from 28 days to under 48 hours
• Bifrost — vDOT liquid staking

Common errors and fixes

• Validator missed blocks. Redelegate to a higher-uptime validator. Use your chain explorer to inspect validator performance metrics before redelegating.
• Slashing event. Native slashing is irreversible. Double-check the validator slashing history before delegating; for liquid staking, the protocol absorbs slashing across the pool.
• Tokens stuck in unstaking. Unstaking has a fixed protocol-level cooldown. Use liquid-staking receipt tokens (e.g. stETH, mSOL, jitoSOL) for instant exit liquidity via DEXes.
• Rewards not accruing. Confirm the delegation transaction succeeded on-chain. Some chains require a separate "claim rewards" transaction; check the chain docs.
• High commission eating yields. Validator commissions can be raised retroactively. Move to a lower-commission operator (typically 3-7%).

FAQ