What Is a Bitcoin Halving?

In brief

A Bitcoin halving is a consensus-enforced reduction in the block subsidy. The subsidy began at 50 BTC per block and steps down by 50% every 210,000 blocks: 50, 25, 12.5, 6.25, 3.125 BTC, and so on until integer division reaches zero.

The block reward is broader than the subsidy:

block reward = block subsidy + transaction fees

Only the subsidy halves. Fees come from the difference between a transaction’s inputs and outputs, so fee income varies with the transactions a miner includes. A halving therefore does not necessarily cut a miner’s total revenue by exactly 50%.

At Bitcoin’s target average of one block per ten minutes, 210,000 blocks is about four years. Actual halving dates vary because individual blocks arrive randomly and the average hash rate changes between difficulty adjustments. No price result follows from the rule.

Subsidy, reward, and coinbase are not synonyms

Bitcoin’s terminology is easy to blur:

• Subsidy: newly created bitcoin that consensus permits at a given block height.
• Transaction fees: existing bitcoin left over when the included transactions’ input values exceed their output values.
• Block reward: the subsidy plus those available fees.
• Coinbase transaction: the first transaction in a block. It has no ordinary inputs and creates outputs up to the permitted subsidy plus the block’s fees.

The coinbase does not move fee outputs from users into a miner’s wallet one by one. Fees are implicit in the included transactions, and consensus allows the coinbase to claim that total together with the subsidy. A miner may claim less than the maximum, but a block that creates more than the permitted subsidy plus fees is invalid and full nodes reject it.

Coinbase outputs cannot be spent until they have 100 confirmations. That maturity rule protects against spending proceeds from a block that later becomes stale; it is not part of the halving schedule.

The trigger is block height

The subsidy era changes when the next block’s height crosses a multiple of 210,000. It does not change at midnight, on an anniversary, or when a countdown website reaches zero.

The first completed eras illustrate the rule:

• heights 0–209,999: 50 BTC;
• heights 210,000–419,999: 25 BTC;
• heights 420,000–629,999: 12.5 BTC;
• heights 630,000–839,999: 6.25 BTC;
• heights 840,000–1,049,999: 3.125 BTC.

Those ranges describe the maximum subsidy per block, not total miner income. They also do not guarantee that every permitted satoshi is created: a miner can underclaim its coinbase, and unclaimed subsidy is not issued later.

Why “about every four years” is only an estimate

The simple estimate is:

210,000 blocks × 10 minutes = 2,100,000 minutes, or about 3.99 years.

Ten minutes is a target average, not an appointment for each block. Bitcoin retargets proof-of-work difficulty every 2,016 blocks toward a two-week interval. The developer guide documents both the 1,209,600-second target and the historical off-by-one detail that measures timestamps spanning 2,015 block intervals.

If blocks arrive faster than the target on average, the next halving arrives earlier than a calendar projection made at the start of the era; slower blocks push it later. Difficulty adjustment pulls the long-run rate back toward the target but does not make a future date certain. A block-height countdown is mechanically relevant; a date countdown is a forecast.

How the finite schedule works

Bitcoin Core calculates how many complete 210,000-block intervals have passed and right-shifts the initial 50 BTC subsidy by that count. Because amounts are whole satoshis, fractional results are truncated. BIP 42 made the zero-subsidy behavior explicit after enough shifts, preventing the old implementation’s undefined shift behavior from ever restarting issuance.

Under the idealized schedule, subsidy issuance totals 20,999,999.9769 BTC, not exactly 21 million. Actual spendable supply is lower because the genesis subsidy is unspendable, some miners have underclaimed rewards, and keys or outputs can become unusable. “21 million” is the conventional rounded cap; it is not a live count of recoverable coins.

The whitepaper’s incentive section says newly issued coins initially motivate block production and that transaction fees can eventually become the incentive after issuance ends. It does not prove that a future fee market will buy any particular amount of hash power. The long-run security budget is an economic question, not a guarantee embedded in the halving formula.

What a halving does—and does not—change

A halving directly changes:

• the maximum new bitcoin created by each valid block;
• the rate at which the theoretical supply ceiling is approached;
• one component of miner revenue.

It does not directly change:

• transaction fees or block-space demand;
• the proof-of-work target or the 2,016-block retarget rule;
• existing balances or UTXOs;
• the maximum block weight;
• the market price;
• the number of miners or the hash rate.

Miners may respond to changed revenue by switching equipment on or off, but that is participant behavior. If hash rate falls, blocks may be slower until difficulty adjusts. Bitcoin does not stop merely because subsidy revenue falls, and a halving itself does not perform the difficulty adjustment.

Halving versus mining and proof of work

The live Bitcoin article gives the whole network overview. The live Proof of work article explains the hash-based consensus mechanism. What Is Bitcoin Mining? focuses on the operational job of building candidate blocks, searching headers, and working through pools. This article has the narrower monetary-policy intent: what quantity changes, at which height, and why subsidy must not be confused with fees.