Bitcoin Halving: The 210,000-Block Rule, Checked on Every Halving Since 2012
A bitcoin halving is the point, every 210,000 blocks, where the number of new bitcoins paid to the miner of each block is cut in half. The fourth came at block 840,000 on 20 Apr 2024 at 00:09 UTC, when the payment fell from 6.25 to 3.125 bitcoin.
Most assets have no published issuance schedule at all. Bitcoin has one written into its software, and the halving is the step in that schedule people talk about. Because the rule is public and every block is recorded, the dates, the amounts and the gaps between halvings can all be read directly from the chain rather than taken on trust. This page does that, then sets out what the price did around the three halvings that have daily price data, and why three cases settle very little.
How it works
Each new block pays its miner newly created bitcoin, called the subsidy. The miner also collects the fees attached to the transactions in that block. The subsidy plus the fees is the block reward.
The subsidy is set by block height, not by date. Bitcoin Core’s GetBlockSubsidy function divides the height by 210,000, starts from 50 bitcoin and halves the amount once for every whole 210,000 blocks already passed. The code comment says this happens “approximately every 4 years”.
The schedule, and the 21 million cap
The schedule so far: 50 bitcoin per block from the first block, 25 from block 210,000, 12.5 from 420,000, 6.25 from 630,000 and 3.125 from 840,000. The next step, to 1.5625, comes at block 1,050,000.
Issuance falls with it. The network aims for one block every ten minutes, or 144 a day. At that pace the 2024 halving cut new supply from about 900 bitcoin a day to about 450.
The cap is the sum of the schedule. Adding up 210,000 blocks at each subsidy, with each halving rounded down to whole satoshis as the code does, gives 20,999,999.9769 bitcoin. That is the precise figure behind the familiar 21 million.
Height, not the calendar
Blocks do not arrive exactly every ten minutes. Mining difficulty adjusts every 2,016 blocks to pull the average back toward ten minutes, but hash power keeps changing in between. So the halvings land on predictable heights and unpredictable dates.
The four gaps, from block explorer timestamps: 1,424.9 days from the first block to the first halving, then 1,319.1, 1,402.1 and 1,439.2 days. At an exact ten minutes a block, 210,000 blocks would take 1,458.3 days. Every era so far ran short of that, because blocks came slightly faster than the target on average.
Where the chain stood on 25 Sep 2026. The latest block was 968,632, mined at 03:27 UTC on 26 Sep, which was the evening of 25 Sep in US time zones. That leaves 81,368 blocks to the fifth halving. Since block 840,000 the average has been 9.95 minutes a block. If that pace held, the remaining blocks would take about 562 days, which would land around 10 Apr 2028; at exactly ten minutes, about 565 days. The height is fixed; the date is an estimate that moves with every block.
A worked example
Look at the two blocks either side of the 2016 halving. Block 419,999, mined at 16:41 UTC on 9 Jul 2016, contained a single transaction: the miner’s own payment of exactly 25 bitcoin, the subsidy with no fees. Block 420,000, mined four minutes later, paid 13.07569681 bitcoin: the new 12.5 subsidy plus 0.57569681 in fees from the 1,256 transactions besides its own payment.
Now the 2024 halving. Block 840,000, mined at 00:09 UTC on 20 Apr 2024, paid its miner 40.75061499 bitcoin. The subsidy was 3.125, so fees made up 37.62561499, about 12 times the subsidy. The block just before it, 839,999, paid 7.29455386: a 6.25 subsidy plus 1.04455386 in fees.
The arithmetic for a miner. Before the 2024 halving, 144 blocks a day paid 144 x 6.25 = 900 bitcoin in subsidy across the whole network. After it, 144 x 3.125 = 450. A miner with a fixed share of the network’s hash power earned half as many new coins the next day, with the same electricity bill, unless fees or the price made up the difference.
The original data
The data: every halving read from mempool.space’s block explorer on 25 Sep 2026, with the subsidy recovered as the block reward minus the fees, and checked against the halving rule in Bitcoin Core’s code. Each block’s hash is in a CSV of the four halvings.
| Halving | Height | Time (UTC) | Subsidy | Fees in that block | Days since previous |
|---|---|---|---|---|---|
| 1 | 210,000 | 28 Nov 2012, 15:24 | 50 to 25 | 13.56295554 | 1,424.9 |
| 2 | 420,000 | 9 Jul 2016, 16:46 | 25 to 12.5 | 0.57569681 | 1,319.1 |
| 3 | 630,000 | 11 May 2020, 19:23 | 12.5 to 6.25 | 0.90968084 | 1,402.1 |
| 4 | 840,000 | 20 Apr 2024, 00:09 | 6.25 to 3.125 | 37.62561499 | 1,439.2 |
The fee column is the part most summaries skip. In 2016 and 2020 the halving block’s fees were under one bitcoin. In 2012 and 2024 they were 13.56 and 37.63 bitcoin, which shows how much the fee side of a miner’s income can swing from one block to the next. The subsidy is the fixed part of the reward; the fees are whatever users are willing to pay at that moment. As the subsidy keeps halving, the share of miner income that depends on fees can only grow.
The price around each halving, from Yahoo Finance’s BTC-USD daily closes. The series starts on 17 Sep 2014, so the 2012 halving has no daily close here and is left out rather than filled from another source. For the other three, the reference is the close of the UTC day the halving block was mined, compared with 365 days before, 180 days after and 365 days after:
- 9 Jul 2016, close $650.96. A year before, $284.89 (a rise of 128.5% into the halving). 180 days after, $1,013.38 (55.7%). A year after, $2,518.44 (286.9%).
- 11 May 2020, close $8,601.80. A year before, $6,972.37 (23.4%). 180 days after, $14,833.75 (72.4%). A year after, $56,704.57 (559.2%).
- 20 Apr 2024, close $64,994.44. A year before, $27,276.91 (138.3%). 180 days after, $67,399.84 (3.7%). A year after, $85,174.30 (31.0%).
Read these as three events, not a rule. All three years after a halving ended higher, but by amounts that differ by a factor of 18, and in 2024 most of the year’s rise came before the halving rather than after it. Each window also contains every other thing that happened to bitcoin and to markets that year, so none of it can be credited to the halving alone. The closes are in a CSV of the price windows. With three observations, the honest summary is that the past gives no reliable guide to what the price does after the next one.
When it fails
As a trading signal. The height of every future halving has been known since the software was written. A supply change that everyone can see years ahead is not new information on the day.
As a four-year calendar. The gaps have ranged from 1,319.1 to 1,439.2 days. A plan pinned to a date rather than a height can be off by months.
When the sample is treated as proof. Three halvings with daily prices, each in a different market, cannot separate the halving’s effect from everything else. The spread of outcomes above is the main finding.
For miners with thin margins. A halving cuts the subsidy overnight while costs stay the same. The mining profit calculator shows how quickly a cut in reward turns a margin negative.
When fees are assumed to fill the gap. The 2024 halving block shows fees can dwarf the subsidy for a moment. The 2016 and 2020 halving blocks show they can also be a small fraction of it.
Related
The bitcoin page covers the asset itself, and mining explains who receives the subsidy each halving cuts. The blockchain page explains the chain of blocks whose heights set the schedule, and buying your first crypto covers sizing a position against falls far larger than a single halving year.
I treat the halving as a known supply change, not a price signal. The date is public years ahead, so whatever it means is already in front of every buyer and seller, and I would not size a position around three past examples.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.