What Is Proof of Work (PoW) in Crypto?

Bitcoin and other proof-of-work chains stay secure only while mining pays, so each network's safety rests on its hash rate, its fee income, and how mining power is spread over time.

Level
Beginner
Reading time
11 min
Last updated

Overview

Introduction

In a proof-of-work (PoW) network, the right to add the next block goes to the first miner to find a block hash below a target number. The only way to find one is to keep guessing, and every guess costs electricity, so each block comes with a price that anyone can check.

Checking the proof takes a full node a single hash, and a block that breaks the rules gets rejected however much power went into it. Online identities are cheap to fake, while hashing work can't be faked at all, which puts a price on any attempt to rewrite the chain.

How Proof of Work Works

Every block starts with users broadcasting signed transactions. Nodes check each one against the basic rules before relaying it to the mempool, the queue of transactions waiting for a block. Miners pick from that queue, assemble a candidate block, set the transaction order, and add a coinbase transaction that pays the block reward to themselves.

From there the race begins. Each miner hashes the block header over and over, changing a nonce and other editable fields between attempts, until one result falls below the network's target. Secure hash outputs can't be predicted, so no miner can skip ahead, and that brute-force search is the core of how crypto mining works.

Once a miner finds a valid hash, it broadcasts the block. Other nodes check the proof and every transaction inside, add the block to their copy of the chain, and the race for the next block starts. If two valid blocks appear at almost the same moment, the chain splits for a while. Nodes settle the split by following whichever branch gathers more accumulated work first, and blocks on the losing branch drop out in a chain reorganization, or reorg.

Below, each step of that cycle is paired with what happens at it.

User signs a transaction

Wallet proves control of the coins being spent

Nodes check basic rules

Bad signatures, bad amounts, and rule-breaking transactions rejected

Miner builds a candidate block

Valid transactions plus a coinbase transaction for the reward

Miner searches for a valid hash

Nonce changed until the header hash falls below the target

Winner broadcasts the block

Every node checks the proof and each transaction

Network extends the chain

Miners build on the valid chain with the most accumulated work

What Proof of Work Does Not Validate

Proof of work decides which block gets the chance to extend the chain. Every transaction inside that block still has to pass the usual checks on signatures, balances, block size, and consensus rules, and the hashing proves only that someone paid to produce the block.

So a miner can choose which transactions go in and propose the block, yet it can't force full nodes to accept anything invalid. A block that creates extra coins, spends someone else's coins, or breaks a protocol rule gets rejected by every node that enforces the rules, and the electricity behind it is lost.

Several common assumptions about mining mix up those two jobs. Each one is set against what the protocol does.

Miners validate everything

Miners propose blocks, and full nodes enforce the rules

Proof of work proves a transaction is legitimate

Signatures and consensus rules prove that, and PoW proves costly block production

The longest chain always wins

The valid chain with the most accumulated work wins

A miner can change any transaction

A miner can order transactions but can’t forge signatures

More nodes means more mining power

Nodes verify and miners hash, two separate roles

Operating a full node lets you check those rules yourself, so you don't have to trust a miner, a mining pool, a wallet provider, or an exchange to tell you which chain is valid.

Hashes, Nonce, and Difficulty Target

Hashing turns the work into something anyone can measure. Bitcoin miners put the block header through the SHA-256 hash function and get a fixed-length number back, and only a result below the network's target counts as valid work. Lowering the target makes the puzzle harder, and raising it makes the puzzle easier.

The nonce is a number in the block header that miners change before each new attempt, and any change to the input produces a completely different hash. Miners try billions of values, and once the nonce field is used up, they change an extra nonce inside the coinbase transaction and keep going. A search with a predictable answer would let an attacker skip the cost, so the repetition is the design.

Hash power joins and leaves the network all the time, which is why Bitcoin retargets every 2,016 blocks. Bitcoin Core, the reference node software, measures how long the last window took, compares it with two weeks, the time 2,016 blocks take at the 10-minute target, and scales the target by that ratio. A window that finished early gets a lower target and a harder puzzle, and a slow one gets an easier puzzle.

Bitcoin Core caps each retarget by treating any window shorter than 3.5 days as 3.5 days and any window longer than 56 days as 56 days. One adjustment can therefore at most quadruple difficulty or cut it to a quarter. A sudden wave of new miners can speed blocks up for a while, and a mass exit can slow them down, until the retargets that follow pull block times back toward 10 minutes.

As of Nov. 16, 2026, the latest retarget came at block 975,744 on Nov. 14, 2026, according to the mempool.space block explorer. The 2,016 blocks before it took about 15.4 days, or close to 11 minutes each, so difficulty fell about 9.1% to roughly 134.8 trillion. The two retargets before that had raised it by about 5.9% and 5.5%.

Bitcoin adjusts mining difficulty every 2,016 blocks

Bitcoin Core scales the proof-of-work target by how long the last 2,016 blocks took against two weeks, and caps each move at four times up or a quarter down.

Faster blocks raise difficulty and slower blocks cut it, so block times drift back toward 10 minutes, though a huge hash rate jump needs more than one retarget.

Hypothetical. Real hash rate changes during a window, and block times vary around their average. Bitcoin Core times the window from its first to its last block, 2,015 intervals, a small skew left out here. The rule shown is Bitcoin's, and other proof-of-work chains retarget differently.

Mining Economics, Block Rewards, and Fee Markets

Proof of work secures a network only as long as miners have a reason to keep hashing. Their revenue comes from the block subsidy, the new coins each block creates, and from the fees users attach to transactions. Their costs include electricity, hardware that wears out and loses value, hosting, and staff.

Bitcoin's subsidy is cut in half every 210,000 blocks on a fixed halving schedule, which slows supply growth and leaves fees to cover a bigger share of security over time. Miners who can't operate near break-even switch off, which removes the least efficient hash power, while those with cheaper electricity or newer machines stay online longer.

Fees count most when blocks are full. Users then bid against each other for space, which gives miners revenue that rises with demand as the subsidy shrinks. The same bidding makes urgent transfers expensive during busy stretches, and PoW never promised cheap transactions in every condition.

That makes PoW security a moving figure, repriced all the time by power costs, the coin's price, and demand for block space. Heavy fee demand can keep miners paid even as issuance falls, and weak fee demand leaves the network leaning on a rising coin price to fund the same hash rate. Whether fees can take over from the subsidy for decades is still an open question.

Security Budget After the 2024 Halving

Block 840,000, mined on Apr. 20, 2024, cut Bitcoin's subsidy from 6.25 bitcoin (BTC) to 3.125 BTC per block. Block 1,050,000, expected around 2028 depending on block times, will cut it to 1.5625 BTC.

The short answers below cover miner pay and the security budget, topic by topic.

What pays miners

Block subsidy plus transaction fees

Fee share of security

Rises with each halving

Effect of low fee revenue

Gradual, cushioned by BTC price, power costs, hardware efficiency, and difficulty

Signals to watch

Hash rate, pool concentration, fee revenue, and demand for block space

Why Proof of Work Is Hard to Attack

The classic PoW threat is a majority attack, often called a 51% attack. An attacker with more hash power than the rest of the network builds a competing chain in private, then releases it once it holds more accumulated work, which reverses recent payments. The attack targets transaction finality. Full nodes would still reject invalid blocks, so the attacker can't create coins from nothing or take coins that never belonged to them.

Section 11 of the Bitcoin white paper by Satoshi Nakamoto works out the odds. An attacker with less than half the hash power sees the chance of catching up drop exponentially with each block it has to replace, and an attacker with more than half catches up eventually.

As of Nov. 16, 2026, the mempool.space explorer estimates Bitcoin's hash rate at about 953 exahashes per second (EH/s). A majority attacker would need more than that, plus the power supply and hardware to sustain it for the whole attack. Other miners keep producing blocks the entire time, so the private chain has to grow faster than a public chain that never stops growing. The larger and more spread out a network's hash power, the higher that bill.

A mining pool lets many miners share rewards, and it also lets one coordinator choose which transactions go into the blocks its members work on. A few large pools therefore mean fewer independent decision-makers. Concentration is a long way from an attack, though it raises censorship and governance concerns if a handful of pools gain outsized influence. So security analysis looks at how hash power is distributed as well as at its total.

PoW security is economic, then, and it holds only while mining by the rules pays better than a sustained attack.

How Blockchain Transaction Confirmations Work

Proof-of-work finality is probabilistic. A transaction gets harder to reverse with every block built on top of it, yet no single block makes reversal impossible.

One confirmation means the transaction is in a block. Each additional one is another block of work an attacker would have to redo and then exceed, so payments with few confirmations face the most risk. You can follow a transaction's position through block height, the count of blocks since the first one. Six confirmations is a common Bitcoin convention, though it's a rule of thumb with no guarantee attached.

What each confirmation count means in practice is set out below.

Zero confirmations

Broadcast but not yet in a block, with the highest double-spend risk

One confirmation

In one block, still close to the chain tip

Two to three confirmations

Often enough for moderate amounts, depending on risk tolerance

Six confirmations

Common Bitcoin benchmark for larger payments and exchange deposits

More than six

Used for very large transfers or unusual network conditions

How long to wait depends on the amount, who's on the other side, and the network. A small payment can tolerate less certainty, while a large exchange deposit, over-the-counter (OTC) trade, or settlement should wait for more blocks.

Energy Use and the Policy Debate

Proof of work draws a lot of electricity because its security comes from computation that never stops. Critics call that an avoidable cost, especially since large proof-of-stake networks have shown lower direct power demand. Supporters answer that the cost is the point, since it ties security to an outside expense that's hard to fake.

One side of the debate tracks total consumption, emissions intensity, and strain on local grids. The other points to where miners set up, their use of curtailed power that would otherwise go to waste, and demand-response programs that pay miners to switch off at peak times. Both sides cite evidence, though they often draw different boundaries and time windows.

The Cambridge Bitcoin Electricity Consumption Index (CBECI), for one, models Bitcoin's power demand from hash rate and mining-hardware efficiency and reports a daily best-guess estimate, so its figures move as those inputs change. A July 2019 commentary from the International Energy Agency (IEA) looked at Bitcoin's electricity use against the wider power system.

In the US, the Energy Information Administration (EIA) estimated in a Feb. 1, 2024, analysis that crypto mining probably accounted for 0.6% to 2.3% of the country's electricity use. The EIA has since discontinued the emergency survey of mining operations behind that tracking.

Energy mix and grid effects vary by region and season, and policy outcomes depend on local power markets, so a single global total settles little on its own. When judging a PoW network, ask where its hash rate is located, which power sources dominate there, and how fast miners can move when rules change.

Where Proof of Work and Proof of Stake Each Fit

Proof of work and proof of stake (PoS) solve the same consensus problem with different costs. PoW miners spend outside resources, hardware and electricity, to produce blocks, and a miner who cheats loses that power and the reward. PoS validators put capital at risk inside the system, as bonded stake that the protocol can slash for misbehavior.

Communities that want security tied to an outside cost and a long operating record tend to pick PoW. Those that put throughput, lower direct power use, and validator access without specialized mining hardware first tend to pick PoS. For any one network, the question to ask is which trust assumptions, failure modes, and governance responses you're willing to accept.

Which Networks Still Use Proof of Work

Bitcoin remains the reference point for proof-of-work security, yet PoW covers many designs. Networks differ in hashing algorithm, block structure, reward model, and miner incentives, and each difference changes the risk profile.

Top PoW Crypto Assets by Market Cap

#CoinPrice24h %MCap24h Vol
1 Bitcoin BTC $62,939.54 -0.03% $1.27T $31.62B
2 Zcash ZEC $922.10 -0.55% $15.65B $669.97M
3 Monero XMR $491.98 -1.4% $9.27B $133.14M
4 Dogecoin DOGE $0.06 -0.25% $9.18B $441.66M
5 Bitcoin Cash BCH $205.19 -0.72% $4.13B $98.93M
6 Litecoin LTC $43.31 -0.9% $3.37B $184.01M
7 Ethereum Classic ETC $5.81 -1.75% $924.67M $25.22M
8 Kaspa KAS $0.03 -0.1% $745.65M $6.28M
9 Dash DASH $36.84 -1.17% $475.21M $43.05M
10 Bitcoin SV BSV $14.92 -0.32% $300M $9.12M

The networks below are listed with their hashing algorithm and what sets each one apart.

Bitcoin

SHA-256, the largest PoW network and the benchmark for hash rate, fee markets, and confirmation practice

Litecoin

Scrypt, 2.5-minute blocks, merge-mined with Dogecoin

Dogecoin

Scrypt, merge-mined with Litecoin through auxiliary proof of work (AuxPoW), so the same work helps secure both chains

Monero

RandomX, privacy-focused, with pool concentration and reorg risk raised by the 2025 Qubic episode

Kaspa

PoW on a blockDAG, a directed acyclic graph (DAG) of blocks ordered by the GHOSTDAG protocol

Bitcoin Cash

SHA-256, mined on the same hardware as Bitcoin, so hash rate can move between the two chains

Zcash

Equihash, privacy-focused, with a security budget and mining spread to judge apart from Bitcoin’s

Chains that share an algorithm share a pool of hardware, so a smaller chain's hash rate can spike or drop as miners from a larger one switch over for a better payout.

How to Evaluate a Proof-of-Work Network

Start with the security model and leave the price chart for later. A PoW coin can have a large market cap and still be cheap to attack if its hash rate is thin, rented, or concentrated in a few pools.

Judging a PoW coin means asking who mines it, what hardware secures it, how much miners earn, how easily hash power moves in or out, and how the network handles reorgs. Each check is named on the left below, with the signal to look for on the right.

Hash rate

Enough active work to make attacks expensive

Mining algorithm

Specialized, widely available, or shared with larger chains

Pool concentration

Share of blocks found by the top few pools

Miner revenue

Split between subsidy and fees

Confirmation practice

Exchange deposit requirements above the norm

Reorg history

Past long reorgs or majority-hash attacks

Developer response

Record of handling security incidents

Proof of work holds up only as well as the economy around it, from miners, fees, hardware, and power access to users who wait for enough confirmations.

FAQs

What is proof of work in blockchain?

It’s a consensus method in which miners must show they spent computing power before adding a block. The idea predates Bitcoin, whose white paper builds on Adam Back’s Hashcash, a proof-of-work system designed as a defense against denial-of-service attacks.

What is proof of work in crypto?

It’s how a network of computers that don’t know each other agrees on one transaction history with no central authority. Bitcoin has used it since its January 2009 launch, and coins such as Litecoin and Monero use the same design with different hashing algorithms.

How does proof of work actually validate transactions?

Strictly, it doesn’t. Signatures and consensus rules validate transactions, and proof of work decides which valid block gets added and in what order. That ordering is what stops the same coins from being spent twice, since the network accepts only the first spend recorded in the chain with the most work.

Why does proof of work use so much energy?

The difficulty adjustment ties the work to the hash rate. When more miners join, the next retarget raises difficulty, so the network still produces about one block every 10 minutes while burning more power to do it. Total energy use follows what miners can afford to spend, which rises with the coin’s price and fees.

Can proof of work be hacked with a 51% attack?

Yes, in principle, and smaller networks face the higher risk. An attacker can rent hash power for the same algorithm for a short window, with no hardware to buy. A chain with little hash rate of its own costs less to overpower than Bitcoin does.

What is proof of work vs. proof of stake?

PoW picks block producers by computing power, and PoS picks them by the coins they’ve staked. Ethereum switched from PoW to PoS in September 2022, and ethereum.org cites an estimate from the Crypto Carbon Ratings Institute (CCRI) that the switch cut the network’s annualized electricity use by more than 99.988%.

Is proof of work the same as mining?

Mining is the activity, and proof of work is the rule it has to satisfy. Bitcoin miners use application-specific integrated circuits (ASICs), chips built only for SHA-256 hashing, and many join pools that combine hash power and split rewards so payouts arrive more steadily.

Do miners validate transactions or do nodes validate them?

Both, at different stages. A miner checks transactions before including them, because one invalid transaction gets the whole block rejected and wastes the work. Full nodes then recheck everything on their own, and only their acceptance makes the block part of the chain.

How many transaction confirmations should I wait for?

Check the exchange’s or recipient’s own requirement first, since many set a fixed confirmation count for each coin. At Bitcoin’s 10-minute target, six confirmations take about an hour on average.

Can you stake a proof-of-work coin?

Not at the protocol level, since a PoW chain has no staking to join. Yield that some exchanges or decentralized finance (DeFi) apps offer on PoW coins comes from lending, wrapped tokens on other chains, or similar arrangements, and it adds counterparty risk without securing the base chain.

What happens to Bitcoin mining after all BTC are mined?

Bitcoin Core’s subsidy rule reaches zero at block 6,930,000, which block-time estimates put around 2140, and from then on fees alone pay miners. If fees fell short and miners left, the difficulty adjustment would lower difficulty to keep blocks coming, at a lower level of security.