TL;DR: Bitcoin has a hard cap of 21 million units, enforced independently by every full node on the network. New bitcoin enter circulation through mining rewards that halve every 210,000 blocks (roughly every four years), following a fixed issuance schedule that runs until the subsidy reaches zero at approximately 2140, just under 21 million. No person, company, or government can change this limit unilaterally. Raising it would take broad, voluntary adoption of new consensus rules by node operators who mostly hold bitcoin themselves, and who would therefore be accepting dilution of their own holdings.
Bitcoin's supply limit is a consensus rule embedded in Bitcoin's source code that caps the total number of bitcoin that will ever exist at just under 21 million. This cap is enforced by every full node on the network independently validating each block against the same rules, rather than by a central authority. Combined with the halving schedule (which reduces new issuance by 50% every ~four years), the supply cap makes bitcoin the first monetary asset in history with a supply that is both verifiably fixed and mathematically enforced.The 21 Million Cap
There will never be more than 21 million bitcoin. More precisely, there will never be more than 20,999,999.9769 bitcoin, because each halving truncates the subsidy to a whole number of satoshis, so issuance stops just short of 21 million.
This number follows directly from the parameters Satoshi Nakamoto chose for Bitcoin's issuance schedule: an initial block reward of 50 bitcoin, halving every 210,000 blocks. The total supply equals 210,000 × 50 × (1 + 1/2 + 1/4 + ...) = 210,000 × 50 × 2 = 21,000,000 BTC. The halving interval is defined by the constant nSubsidyHalvingInterval = 210000 in Bitcoin Core's src/kernel/chainparams.cpp.
The specific number matters less than the principle behind it: a fixed, finite supply that every participant can verify independently and no one can alter unilaterally. Whether the cap were 21 million or 100 million, the essential property is the same. What matters is that the cap exists, that it is known, and that it is enforced.
This article explains the cap mechanism itself: the issuance schedule, the halvings, and how every node enforces the limit. For the broader economic argument, why a fixed, rules-based issuance makes for better money than discretionary central-bank policy, see why Bitcoin offers better monetary policy.
As of 2026, approximately 20 million bitcoin have already been mined. That means roughly 95% of all bitcoin that will ever exist are already in circulation. The remaining 1 million will be distributed to miners over the next 114 years, in steadily shrinking amounts.
How Much Bitcoin Is Left to Mine?
About 1 million bitcoin remain to be mined as of 2026, roughly 5% of the 21 million cap. Because the block reward halves every four years, this remaining supply enters circulation in ever-shrinking amounts: more than 90% of the bitcoin still to be mined will be issued before 2040, yet the last fraction of a bitcoin will not be mined until around 2140. The pace slows even as the count creeps toward the cap.
The Issuance Schedule: How New Bitcoin Enter Circulation
New bitcoin do not appear all at once. They enter circulation through a process called mining, where miners compete to add new blocks to the blockchain. Each time a miner successfully produces a valid block, the protocol rewards them with a specific amount of newly created bitcoin. This is the block subsidy; together with the block's transaction fees it makes up the miner's total block reward.
When Bitcoin launched in January 2009, the block reward was 50 bitcoin per block. Every 210,000 blocks (approximately every four years), the reward is cut in half. This event is called the halving.
| Event | Approximate Date | Block Height | Block Subsidy | New Bitcoin per Day (~144 blocks) |
|---|---|---|---|---|
| Genesis (launch) | January 3, 2009 | 0 | 50 BTC | ~7,200 BTC |
| First halving | November 28, 2012 | 210,000 | 25 BTC | ~3,600 BTC |
| Second halving | July 9, 2016 | 420,000 | 12.5 BTC | ~1,800 BTC |
| Third halving | May 11, 2020 | 630,000 | 6.25 BTC | ~900 BTC |
| Fourth halving (most recent) | April 19, 2024 | 840,000 | 3.125 BTC | ~450 BTC |
| Fifth halving (projected) | ~2028 | 1,050,000 | 1.5625 BTC | ~225 BTC |
The pattern continues: each halving cuts the reward in half again. After 33 halvings (around the year 2140), the reward reaches 0 satoshis because the subsidy calculation uses integer right-shift (nSubsidy >>= halvings) which truncates fractional satoshis. At that point, no new bitcoin will be created. The issuance is complete.
This schedule is deterministic. Every halving is already defined in the code, so any participant can compute the issuance at any past or future block height, including in 2140. No other widely held monetary asset publishes its future issuance to this precision.
How the Cap Is Enforced
A supply cap is only meaningful if it can be enforced. The dollar has no supply cap because the Federal Reserve can create new dollars at will. Gold has a natural supply cap (the total amount in the Earth's crust), but no one can verify exactly how much gold exists or will be discovered. Bitcoin's supply cap is enforced by a mechanism that has no equivalent in traditional monetary systems: decentralized consensus.
Full Nodes Independently Validate Every Block
A Bitcoin node is a computer running Bitcoin software that independently verifies every transaction and every block against the protocol's rules. One of those rules is the block reward amount. If a miner creates a block that claims a subsidy of 6.25 bitcoin when the current valid subsidy is 3.125, every full node on the network rejects that block as invalid. It does not matter how much hashrate the miner controls. A block that violates the consensus rules is treated as if it does not exist.
Under the Hood: How the Code Enforces the Cap
What caps the total is the subsidy calculation rather than any single ceiling constant. GetBlockSubsidy() in src/validation.cpp derives the correct subsidy for a block height by halving the initial 50 BTC once for each completed 210,000-block interval, and every full node rejects a block whose coinbase claims more than that amount plus the block's fees. Summed across every height, that calculation is what produces the 21 million figure. A separate constant, MAX_MONEY in src/consensus/amount.h, sets 2,100,000,000,000,000 satoshis as the largest valid amount; Bitcoin Core's own comment describes it as a sanity check rather than the limit on issuance.
Anyone running a Bitcoin node can check the issuance for themselves. The gettxoutsetinfo RPC command sums every unspent transaction output in the UTXO set and reports that total to the satoshi. The figure is the amount currently sitting in spendable outputs, which is slightly below what the schedule allowed: subsidies some early miners never fully claimed, and coins sent to provably unspendable outputs, are not in the set.
Enforcement needs no human arbiter: no committee decides whether a block is valid, and there is no appeals process. The rules are encoded in software, and nodes around the world apply them independently and automatically. A miner who tries to award themselves extra bitcoin wastes the electricity they spent producing the block, because no one will accept it.
Why the Cap Cannot Be Changed
Changing the 21 million cap would require changing Bitcoin's consensus rules, which means persuading the operators of a large, independently run node population to install software that accepts a different supply schedule. That is a coordination problem with no central lever: no authority can push the change through, and any operator who declines simply keeps validating under the existing rules. Node operators who hold bitcoin would also be accepting dilution of their own holdings, so the incentive runs against adoption. No such change has been adopted in 17 years of operation.
This stands in contrast to fiat monetary systems, where the people who decide to create new money are different from the people whose savings are diluted by it. Bitcoin aligns incentives: the people who enforce the rules are the same people who benefit from the rules being enforced.
The Role of Miners vs. Nodes
Miners produce blocks, and nodes validate them. That division of labor is what makes the cap self-enforcing. A miner can include any data they want in a block. But if the block violates any consensus rule, including the supply cap, nodes will reject it. Miners who produce invalid blocks earn nothing. For a miner, violating the supply cap is economically identical to not mining at all.
Why Satoshi Chose a Fixed Supply
Satoshi Nakamoto never published a detailed economic treatise explaining the choice. But the reasoning is embedded in the design itself, and Satoshi's early writings provide direct evidence of the thinking.
The Problem Bitcoin Was Built to Solve
The genesis block (block 0, mined on January 3, 2009) contains a message Satoshi embedded in the coinbase transaction: "The Times 03/Jan/2009 Chancellor on brink of second bailout for banks". This headline from The Times of London references the 2008 financial crisis, during which governments created trillions of dollars, euros, and pounds to bail out failing banks. The savers and workers whose purchasing power was diluted had no say in the decision.
Satoshi wrote in a February 2009 forum post: "The root problem with conventional currency is all the trust that's required to make it work. The central bank must be trusted not to debase the currency, but the history of fiat currencies is full of breaches of that trust."
A fixed supply narrows this trust requirement. No central authority decides the issuance rate, and changing that rate would require the network's participants to coordinate on adopting new rules rather than one institution deciding to.
The Gold Analogy
Satoshi drew explicitly on gold as a model. In a 2010 forum post, Satoshi wrote: "As a thought experiment, imagine there was a base metal as scarce as gold but with the following properties: boring grey in colour, not a good conductor of electricity, not particularly strong... not useful for any practical or ornamental purpose... and one special, magical property: can be transported over a communications channel."
Gold served as money for thousands of years because it had the properties of sound money: it was scarce, durable, divisible, fungible, and difficult to counterfeit. Its scarcity was enforced by geology. Bitcoin replicates these properties digitally, with scarcity enforced by mathematics and decentralized consensus rather than the physical limits of the Earth's crust.
The Problem With Inflation: Purchasing Power Erosion
When the money supply grows faster than the economy produces goods and services, and money demand does not absorb the difference, each unit of currency can buy less over time. For anyone holding cash, that fall in purchasing power has the same effect on their savings as a small annual charge.
The Numbers
Between 2020 and 2022, the U.S. M2 money supply expanded by more than 40%, from approximately $15.4 trillion to over $21.7 trillion. That roughly $6 trillion is growth in broad money, most of which is created when commercial banks lend, rather than dollars the Federal Reserve issued directly; the Fed's own balance sheet grew by a smaller amount over the same period, from about $4.2 trillion to about $8.9 trillion. Consumer price inflation reached its highest level in four decades during this stretch, peaking at 9.1% year-over-year in June 2022. How much of that was driven by money growth as against pandemic supply disruption and fiscal transfers is still debated among economists.
The effects compound. A dollar in January 2000 had the purchasing power of approximately $0.53 by January 2025, according to the Bureau of Labor Statistics' CPI calculator. A person who held $100,000 in cash for 25 years did not lose any dollars, but their savings bought roughly half as much.
Low positive inflation is an explicit policy target rather than an accident. The Federal Reserve aims for 2% per year, on the view that a small buffer above zero guards against deflation and leaves policy room to act. The arithmetic still compounds: at a sustained 2%, purchasing power falls by roughly 45% over 30 years, and at 3%, by roughly 59%.
Why Bitcoin's Fixed Supply Is Different
Bitcoin's issuance rate is currently approximately 0.85% per year (post-2024 halving) and declining. After the 2028 halving, it will drop to approximately 0.4%, falling below 0.2% after 2032 and becoming negligible by 2040. No committee, emergency, or election can change this rate.
New bitcoin still enter circulation, but no one can increase the supply beyond the schedule the protocol already fixed. If demand grows while the supply schedule holds, the market price of each unit may rise. Nothing in the protocol guarantees it, because the schedule governs supply and says nothing about demand.
Bitcoin's Supply Compared to Fiat Money and Gold
Every monetary system has rules (or the absence of rules) governing how new units enter circulation. Comparing the three dominant models illustrates what makes Bitcoin's approach distinct.
| Property | U.S. Dollar (Fiat) | Gold | Bitcoin |
|---|---|---|---|
| Total supply cap | None | ~216,000 metric tons mined (as of 2024, per World Gold Council); unknown remaining in Earth's crust | 21 million bitcoin (hard cap) |
| New supply rate | Discretionary (central bank) | ~1-2% per year (variable) | Fixed by code, halves every ~4 years |
| Supply predictability | Unknown beyond next policy meeting | Roughly predictable; surprises possible | Known precisely through 2140 |
| Who controls supply | Central bankers (small committee) | Geology + mining economics | No central issuer; rules enforced through decentralized consensus |
| Current annual issuance rate | Varies widely (often 2-10%+) | ~1.5% | ~0.85% (post-2024 halving) |
| Supply verifiability | Requires trusting government reports | Estimated; cannot be audited with comparable precision | Anyone running a full node can recompute the issuance schedule and audit the UTXO-set total |
| Can supply rules change? | Yes, at any committee meeting | New discoveries or technology can alter supply | Requires broad, voluntary adoption of new rules across independently operated nodes |
Fiat: Unlimited, Discretionary
The U.S. dollar has no supply cap. The Federal Reserve influences monetary conditions through policy rates, reserve settings, and asset purchases, with no predetermined quantitative limit and no fixed schedule, and its stance can change from one meeting to the next. Most broad money is created when commercial banks lend, so the total reflects bank and borrower behaviour alongside central bank policy. The people setting that policy work under political pressure and imperfect information.
Gold: Scarce but Uncertain
Gold is scarce, and no one can print more of it. Annual mine production adds roughly 1% to 2% to the existing above-ground supply. But the exact rate depends on geology, mining technology, energy costs, and the price of gold itself. New discoveries, improved extraction techniques, or rising prices that make marginal deposits profitable can all increase the supply rate unpredictably. Asteroid mining remains speculative and is sometimes raised as a possible future supply source.
Gold's total supply also cannot be audited with the same precision. No one can state exactly how much gold sits above ground, in private vaults, or remaining in the Earth's crust. The published totals are estimates.
Bitcoin: Scarce, Known, Verifiable
Bitcoin's issuance rules are published, and anyone running a full node can recompute the schedule and audit the UTXO-set total for themselves rather than relying on a figure someone else reports. The future schedule is fixed in the protocol, and the rules governing it are enforced through decentralized consensus rather than institutional authority. Changing them would require the network to adopt new rules.
What Happens When All Bitcoin Are Mined
Around the year 2140, the final halving will cut the block reward from one satoshi to zero. No new bitcoin will be created after that point. Miners will continue operating, but their compensation will come entirely from transaction fees.
The Fee-Based Security Model
Miner revenue already has two components: the block reward (newly created bitcoin) and transaction fees (paid by users who want their transactions included in a block). Today, the block reward dominates. But with each halving, the balance shifts incrementally toward fees.
This is a 131-year transition, not a sudden switch. By the time the block reward reaches zero, it will have been economically negligible for decades. Consider the trajectory:
- By 2032, about 98.4% of all bitcoin will have been mined. The block reward will be less than one bitcoin per block.
- By 2048, the block reward will be less than 0.05 bitcoin per block.
- By 2060, the block reward will be measured in hundreds of thousands of satoshis, not in whole bitcoin.
Because the subsidy halves on a fixed schedule while fee revenue depends on demand for block space, the subsidy's share of miner revenue declines mechanically and becomes negligible well before 2140. Whether fees grow enough to sustain security spending at current levels is an open question and an active area of research.
Where Do the Fees Come From?
Transaction fees are paid by users who want their transactions confirmed on the blockchain. The fee is market-driven: during periods of high demand for block space, users compete by offering higher fees, and miners prioritize the highest-paying transactions. Sources of block space demand include:
- On-chain bitcoin transactions (sending bitcoin between addresses)
- Layer-2 settlement transactions (Lightning Network channel opens and closes, Liquid Network peg-ins and peg-outs)
- Data anchoring and timestamping
- Any future use of Bitcoin block space that users value enough to pay for
On the day of the April 2024 halving, transaction fees temporarily exceeded the block reward due to a surge in on-chain activity. This demonstrated that fee revenue can be substantial during high-demand periods, even with the block reward still active.
The Difficulty Adjustment as a Safety Net
If transaction fees ever prove insufficient to compensate miners at current hashrate levels, some miners will shut down. When hashrate drops, Bitcoin's difficulty adjustment (which recalculates every 2,016 blocks) lowers the work required per block, which reduces the cost of producing one and improves the margin for the miners who remain. Whether any given miner is profitable still depends on their energy costs and on the bitcoin price. The mechanism targets the block interval rather than miner profits: the network continues producing blocks that average roughly ten minutes over each retarget period rather than halting.
The Smallest Unit: Satoshis and the Divisibility of Bitcoin
One bitcoin is divisible into 100 million smaller units called satoshis (sats). This means the 21 million bitcoin supply is equivalent to 2.1 quadrillion satoshis (2,100,000,000,000,000 sats).
Why Divisibility Matters
A common objection to Bitcoin's fixed supply is: "21 million units aren't enough for eight billion people." This treats bitcoin as if it were indivisible, like a gold bar that cannot be split. It misunderstands how the system works.
If Bitcoin were adopted globally and each bitcoin were worth millions of dollars, everyday transactions would be denominated in satoshis, not whole bitcoin. A cup of coffee might cost 500 sats. A monthly salary might be 200,000 sats. The math works the same way it works with any currency: the unit size adjusts to the price level.
At 2.1 quadrillion total units, there are approximately 262,500 satoshis for every person on Earth. Whether a monetary system can serve a global population depends on how small its smallest unit is relative to the price level, which is a question about divisibility rather than about the count of whole bitcoin.
Denominations in Practice
The Bitcoin community already uses satoshi-denominated pricing. Common conventions include:
- 1 sat = 0.00000001 bitcoin (the smallest on-chain unit)
- 1,000 sats = 0.00001 bitcoin
- 100,000 sats = 0.001 bitcoin
- 1 million sats = 0.01 bitcoin
- 100 million sats = 1 bitcoin
Wallet software, including the Blockstream app, lets users display balances in either bitcoin or sats. As the value of one bitcoin rises, satoshi denomination becomes the more practical daily unit, the same way people think in cents rather than whole dollars for small amounts.
Can Bitcoin Be Made More Divisible?
On the base layer, one satoshi is currently the smallest unit. The Lightning Network tracks channel balances and routes payments in millisatoshis (thousandths of a satoshi), which makes micropayments practical off-chain, though those balances settle on-chain in whole satoshis. If future demand required further subdivision on the base layer, the protocol could theoretically be updated. But with 2.1 quadrillion satoshis available, the need for additional divisibility is not a foreseeable concern.
"21 Million Isn't Enough for the Whole World"
This is the most frequently repeated objection to Bitcoin's fixed supply, and it reflects a misunderstanding of what a supply cap means in a divisible system.
The Scarcity Objection
The argument goes like this: there are roughly eight billion people on Earth and only 21 million bitcoin. If every person on the planet wanted some, there would not be enough to go around. Therefore, Bitcoin cannot function as a global monetary system.
Why the Objection Fails
The objection confuses units with utility. Gold served as a global monetary base for millennia despite a small per-person share (its ~216,000 metric tons work out to about one ounce each), because it was divisible enough that any amount could be represented and exchanged. Bitcoin is far more divisible still: as the divisibility section above shows, 21 million bitcoin is 2.1 quadrillion satoshis, roughly 262,500 satoshis per person. A monetary system needs enough divisibility to represent any amount accurately rather than a large number of base units.
Lost Bitcoin and Effective Supply
Some bitcoin is unlikely ever to move again. Published estimates of the permanently lost portion commonly land between three and four million bitcoin, covering forgotten recovery phrases, discarded hardware, and deaths without succession plans, and usually counting the roughly one million early-mined bitcoin attributed to Satoshi Nakamoto that have never been spent. These figures are estimates by construction: a coin that has not moved cannot be distinguished on-chain from a coin whose owner is simply waiting, so the true total is unknowable.
Coins that never move reduce the supply available to trade. If four million bitcoin were permanently inaccessible, the tradable total would be nearer 17 million than 21 million, leaving the same demand competing for fewer available units. The conditional matters, because the four million is an estimate rather than a measurement.
Lost bitcoin cannot be recovered, re-minted, or replaced. There is no "lost and found" process. Permanent loss reinforces the importance of secure self-custody practices, including proper backup of recovery phrases and the use of hardware wallets like Jade Plus ($149-$169) that keep private keys offline and under the owner's control. For those just getting started, the Jade Classic ($79) provides entry-level hardware wallet security with the same self-custodial model.
Frequently Asked Questions
Why did Satoshi choose 21 million as the cap?
The 21 million cap is a mathematical result of Bitcoin's issuance parameters: an initial block reward of 50 bitcoin, halving every 210,000 blocks. The sum of this geometric series (210,000 × 50 × 2) equals exactly 21,000,000. Satoshi described the choice as an educated guess in April 2009 correspondence, explaining that the aim was to land unit prices in a range comparable to existing currencies without knowing how widely the network would end up being used. The specific number matters less than the property it produces: a fixed cap that participants can verify and that no one can raise unilaterally. The constant MAX_MONEY in Bitcoin Core's source code enforces this cap as a consensus sanity check on every transaction.
Can the 21 million cap ever be increased?
Changing the cap would require altering Bitcoin's consensus rules, which means convincing the vast majority of node operators, miners, exchanges, and users to install new software. Since most of these participants hold bitcoin, they would be agreeing to dilute the value of their own holdings. Their financial interest runs against it, and no such change has been adopted in 17 years of Bitcoin's operation.
What happens when all 21 million bitcoin are mined?
When the block reward reaches zero (projected around 2140), miners will be compensated entirely through transaction fees paid by users. This is by design: the block reward bootstraps the network's security, and transaction fees sustain it long-term. The transition happens gradually over more than a century, with each halving shifting the balance incrementally from block rewards to fees.
Are 21 million bitcoin really enough for the whole world?
Each bitcoin is divisible into 100 million satoshis, making the total supply 2.1 quadrillion sats. Distributed across eight billion people, each person would hold 262,500 sats. The number of whole bitcoin is small, but the number of usable units is enormous. A global monetary system needs sufficient divisibility, not a large number of base units, and Bitcoin provides that in abundance.
How many bitcoin have been lost forever?
Nobody knows, and the figure is not directly measurable: an unmoved coin looks the same on-chain whether it is lost or simply held. Published estimates commonly range between three and four million bitcoin, covering lost recovery phrases, discarded hardware, and unreachable early-miner wallets, and usually including the roughly one million early-mined bitcoin attributed to Satoshi that have never been spent. To whatever extent those coins are genuinely gone, they reduce the supply available to trade.
Does Bitcoin's fixed supply make it deflationary?
Bitcoin's monetary supply has a declining issuance rate and a hard cap, which means its supply-side inflation approaches zero over time. Whether bitcoin is deflationary in terms of purchasing power depends on demand. If demand grows while the supply schedule holds, the market price of each unit may rise, though nothing in the protocol guarantees it. Fiat currencies are instead managed toward a small positive inflation rate (central banks target roughly 2%).