The Difficulty Adjustment: The Heart of Bitcoin

The difficulty adjustment determines how hard it is for miners to find a new Bitcoin block. Every 2’016 blocks, roughly every two weeks, every Bitcoin node compares the actual time with the planned time. If blocks came too fast, it gets harder; if they came too slowly, easier. That keeps the pace at about one block every 10 minutes.

Key takeaways

• The goal: one block every 10 minutes, no matter how many miners are currently hashing.
• Every 2’016 blocks, each node calculates the new difficulty itself. Nobody votes, nobody decides.
• An adjustment is at most ×4 or ÷4. The largest drop since 2010: −27.94% in July 2021, after China's mining ban.
• Without this rule, new bitcoin would arrive sometimes faster, sometimes slower. With it, halving dates and the 21 million limit stay predictable.

Why Bitcoin needs a difficulty adjustment

Imagine the number of miners doubles overnight. Without a counter-rule, blocks would come twice as fast, and new bitcoin would be issued twice as fast. If half the miners drop out, the network stalls and transfers wait twice as long.

Bitcoin has no boss who steps in when that happens. Satoshi Nakamoto described the problem in the white paper: because hardware gets faster and the number of miners varies, the difficulty should follow the average number of blocks per hour. If blocks are found too quickly, it rises.

That is exactly what the difficulty adjustment does. It is the network's pacemaker: computing power may grow a thousandfold, but the heartbeat of about one block every 10 minutes stays.

How the difficulty adjustment works

Miners search for a number that puts the fingerprint (hash) of their block below a target value. This only works by trial and error. The smaller the target, the more attempts it takes on average. The difficulty says how much harder this is today than for the first block in January 2009 (difficulty 1).

The adjustment runs in four steps:

1. The nodes count off 2’016 blocks. At 10 minutes per block that would be exactly 14 days.
2. They measure how long these blocks really took, using the timestamps in the blocks.
3. They divide the planned 14 days by the measured time. A result of 1.05 means 5% harder.
4. If the ratio is greater than 4 or smaller than ¼, it is capped at these limits. The new difficulty applies from the next block.

Every node checks this for itself. A block with the wrong difficulty is rejected by all the others. That is why no miner and no company can get around the rule.

The next difficulty adjustment, live

We fetch these values from the newest blocks every two minutes and calculate them using Bitcoin's own rule. How we do it is in the methodology.

On Nov 12, 2026, 1’851 of the 2’016 blocks had been found. They came on average every 10:52 minutes, which is too slow. If it continues like this, the difficulty will fall by about 8% around Nov 14, 2026 at block 975’744. The previous adjustment on Oct 29, 2026 was +5.54%.

What today's difficulty means in numbers

Since Oct 29, 2026 the difficulty has been about 148.4 trillion (148.4 T). On average the network therefore needs about 640 sextillion attempts per block. The all-time high was 156.0 T on Oct 29, 2025.

From the difficulty and the block time you can estimate the computing power of all miners, the hashrate. mempool.space estimated it at about 936 EH/s on Nov 12, 2026, that is 936 quintillion attempts per second. For comparison: at the end of 2015 the difficulty was 0.1 T, about 1’400 times lower.

Over the years the direction is clear: of 468 adjustments up to the end of October 2026, 358 went up and 110 went down. New, more efficient machines are added, and more miners want a share of the block reward.

The largest difficulty adjustments in history

The steepest drops show when a lot of computing power disappeared at once. All values from mempool.space block data:

In early summer 2021, about half the hashrate disappeared within a few weeks. In the period before the record, blocks came on average only after almost 14 minutes, and the network kept running anyway. After the adjustment the pace was back at 10 minutes. Miners rebuilt their machines in the US, Kazakhstan and elsewhere.

In 2026 too the difficulty swings noticeably: −11.16% on Feb 7, 2026, then +14.73% on Feb 19, 2026 and −10.09% on Jun 14, 2026. For bitcoin holders that changed nothing; for miners it changed a lot.

Worked example: what an adjustment means for a miner

Take a miner with 200 TH/s, roughly a current machine. With a network hashrate of 936 EH/s, it has about 0.00002% of the computing power. Since the 2024 halving each block pays a 3.125 BTC subsidy, plus fees, which we leave out here.

Prices: €64’983 and CHF 60’056 per bitcoin (mempool.space, Nov 12, 2026). After the adjustment the miner earns about 8.6% more without doing anything. This is how the network pulls miners back into the game when others give up.

Conversely, every increase eats part of the miner's income. If the difficulty rises by 30% over a year, the same machine gets about 23% fewer sats. Electricity and the machine still cost the same, though.

Whether such a machine can even pay its way on German or Swiss household electricity is shown in Bitcoin mining at home: is it worth it? Run the numbers with your own values for the machine, electricity price and bitcoin price:

The difficulty adjustment, the halving and the 21 million

Bitcoin only issues new coins through the block reward, and it halves every 210’000 blocks. How quickly the 21 million are reached therefore depends on the number of blocks, not the number of days.

Only the difficulty adjustment turns the block count into a schedule. At 10 minutes per block, 210’000 blocks take about four years. The next halving at block 1’050’000 is therefore roughly in spring 2028. Because the hashrate usually grows and the adjustment lags behind, blocks have come slightly faster on average since 2009, every 9.6 minutes. The halvings therefore each came a few months earlier than after exactly four years.

The computing power gathered behind the difficulty is also what protects the network. Anyone who wants to rewrite the blockchain would need more computing power than all honest miners combined. At 936 EH/s that means machines and electricity on the scale of entire countries.

Limits of the difficulty adjustment

It lags behind. The rule only looks at the last two weeks. If the hashrate suddenly falls sharply, blocks stay slow until the next adjustment. In June 2021 the period therefore took almost three weeks instead of two.

The timestamps are not exact. Miners enter the time in the block themselves, within narrow limits. This leaves a theoretical gap, the so-called timewarp attack, which would be possible if a majority of miners worked together. The proposal BIP 54 ("Consensus Cleanup") is meant to close it.

It knows only one goal. The adjustment keeps the pace. Whether mining pays off is decided by the price, the electricity price and the efficiency of the machines, not by the rule.

What the difficulty means for you as a bitcoin holder

For you as a long-term holder, the difficulty adjustment is above all a reason to stay calm. Whether miners pile in by the thousands or flee: new bitcoin arrive at the planned pace, and there will never be more than 21 million.

The difficulty is not a buy signal. It follows the price with a delay, because miners add machines when prices rise and switch them off when prices fall. Where the price currently stands relative to its long-term trend is shown in the Power Law analysis with floor and zones.

How mining works overall, what miners earn and why their work protects the network is explained in the overview Bitcoin mining explained.

If you want to go deeper: all articles on mining, energy and technology are in Mining & Technology. The basics of blocks and miners are explained in stage 1 of the Bitcoin Academy. Whether your own miner is worth it as a lottery ticket is shown in the article Solo mining: the Bitcoin lottery.