Every smart-money video talks about order blocks. Almost none of them tell you how often price actually comes back to one, or what happens when it does. This lesson does both — the concept in plain language, and the measurement underneath it.
An order block is the last candle that closed against the direction of a strong move, marked as a zone because that is where the buying or selling that caused the move was quietly happening.
An order block is the last candle that closed against the direction of a strong move, immediately before that move began.
For a move up, it is the last red candle before price ran higher and broke a previous high. For a move down, the last green candle before price collapsed. That candle's high-to-low range becomes a zone drawn on the chart, and the claim is that price often comes back to that zone before continuing.
Notice what makes the marked candle special. It is not the biggest candle, and it is not the one you would circle if you were looking for something dramatic. It is the quiet one immediately before the dramatic part. That is the whole idea.
Imagine a big buyer who wants ten thousand shares but knows that buying all ten thousand at once will push the price up against them. So they buy quietly while the price is still drifting down, absorbing what sellers offer, leaving the chart looking weak. Then, once they have most of what they wanted, they stop being careful and take everything available — and the price jumps.
The quiet candle just before the jump is where the careful buying happened. If that buyer has more to do, the place they were willing to buy before is the place they are willing to buy again. That is the reasoning behind the zone. Whether it holds up is a separate question, and one this page answers with numbers rather than a story.
Here is the part almost nobody publishes. Every order block in five years of hourly candles was located by rule, then followed forward to see what happened.
| Instrument | Order blocks found | Price came back | Median wait | Of those, held | Of those, broke |
|---|---|---|---|---|---|
| XAUUSD | 1,158 | 66.6% | 3 h | 80.3% | 19.1% |
| GBPUSD | 1,070 | 70.4% | 3 h | 81.1% | 18.1% |
| EURUSD | 1,048 | 68.8% | 3 h | 74.8% | 23.9% |
| US30 | 713 | 67% | 3 h | 75.5% | 22.2% |
| NAS100 | 753 | 66.9% | 3 h | 74.6% | 24.4% |
What was counted: The last down-closing H1 candle before a close above the most recent swing high. Revisited means price traded back to the top of that candle within 48 hours; held means it then reclaimed the break level without closing below the bottom of the candle. Measured on H1 candles across five instruments, 2021-01-03 to 2026-08-30.
A zone "held" if price reclaimed the level that was broken without first closing below the bottom of the block. It "broke" if it closed through the bottom. The remainder had done neither within 48 hours.
Three things worth taking from that table.
"Held four times out of five" is not "wins four times out of five". Holding here means price reclaimed the broken level before closing below the block — it says nothing about where your stop was, what you paid in spread, or whether you would have sat through the drawdown along the way. It measures the zone, not a trade. Turning one into the other is your job, and the only way to find out is to test your own version.
Aisha trades gold on the 15-minute chart and marks her levels from the hourly.
What she sees. Gold drifts down through the morning, then at 13:00 a single strong hourly candle closes above the high made two hours earlier. That is her break.
What she marks. She walks back from the breaking candle. The 12:00 candle closed green. The 11:00 candle closed red — that is the last down-close before the move. She draws a box from its high to its low and extends it right.
What she does not do. She does not buy the breakout. Price is now well above her box, and buying here means her stop has to sit below the whole zone — a wide stop on a position taken at the worst price of the move.
What she waits for. Price pulls back over the next two hours and touches the top of the box. Now she has a decision point with a natural invalidation: below the bottom of the box, the idea is simply wrong.
How she sizes it. Her stop sits a little under the box. She measures that distance in pips and checks it against the Stop Loss Reality Checker — if the box is so tight that her stop lands inside gold's ordinary noise, she either widens it and trades smaller, or she skips the trade. The box being pretty does not make the stop survivable.
Reading about order blocks costs nothing and teaches almost nothing. Replaying them does. Load gold or GBPUSD in the free backtester, step forward bar by bar, and every time you see a break of structure, mark the last opposing candle and write down what happened next. Thirty repetitions and you will spot them without thinking.
Practise this in the free backtester →Start the drill in the free backtester → No sign-in needed to begin.
An order block is one answer to "where do I get in". It only makes sense after you can see a break of structure, because the break is what identifies the block. It is often paired with a fair value gap, since the same violent move that breaks structure usually leaves an imbalance behind it — and when a block and a gap overlap, you have two reasons for one price instead of one.
It is also worth knowing what an order block is not. It is not a support level, it is not a magic zone, and it is not a reason to skip position sizing. The numbers above are encouraging by the standards of this series, and they still describe a thing that fails one time in five.
What is an order block in simple terms?
It is the last candle that closed the opposite way to a big move, right before that move started. Going up, it is the last red candle before the surge. You draw a box around that candle, and the idea is that price often returns to it before continuing.
Do order blocks actually work?
Under the stated definition, on hourly gold candles, price returned to the block 66.6% of the time within two days, and when it did, the zone held 80.3% of the time. Those are real numbers on a sample of 1,158. But "the zone held" is not the same as "a trade made money" — that depends on your stop, your target and your costs, which only your own testing can tell you.
Should I use the candle body or the full wick?
The measurement here uses the full high-to-low range, which is the more forgiving choice — a body-only block is smaller, so it gets touched less often but offers a tighter stop. Neither is correct in the abstract. Pick one, use it consistently for fifty trades, and let your own results decide.
What timeframe should I draw order blocks on?
Hourly or four-hourly for most people, then execute on a lower timeframe. Below fifteen minutes the "structure" being broken is often noise, and you end up with dozens of blocks that mean nothing. These measurements were taken on hourly candles for exactly that reason.
How is an order block different from support and resistance?
Support and resistance is a level where price previously turned. An order block is a zone defined by what happened immediately before a break — it is about the origin of a move rather than the memory of a bounce. They often land in similar places, and when they do, that overlap is worth more than either alone.
What invalidates an order block?
A close through the far side of the box. If price closes below the bottom of a bullish block, the reason for the zone is gone and holding on is hope rather than analysis. In the sample above that happened about one time in five.
Replay gold, GBPUSD or the Dow bar by bar, mark order blocks in real time before you know the outcome, and record what happened. That is how the pattern goes from something you read about to something you can see.
Open the free backtester →Beginner exploration
Open each answer for a plain-language way to read Order Blocks, Explained Properly, test it carefully and decide what to explore next.
This page focuses on “Order Blocks, Explained Properly”.What an order block is, how to find one, and what happened to 1,158 of them on gold: 66.6% were revisited and 80.3% of those held. Beginner-friendly, with diagrams and a practice drill.For “Order Blocks, Explained Properly”, a beginner should identify what the learning guide measures, assumes or teaches before acting on its conclusion.Treat this page's account of “Order Blocks, Explained Properly” as a learning reference rather than a prediction, signal or promise of future performance.
For “Order Blocks, Explained Properly”, translate the idea into a definition you could apply the same way on two different charts.While exploring “Order Blocks, Explained Properly”, work through one example slowly and record which inputs or observations determined the result.Keep your “Order Blocks, Explained Properly” record honest: list the limitation or counterexample before using the concept in a trading plan.Before leaving “Order Blocks, Explained Properly”, practise the definition on unseen history and review consistency before judging performance.
Turn one idea from “Order Blocks, Explained Properly” into a rule with explicit inputs, dates, costs and pass-or-fail conditions.Ask AI to expose missing assumptions in that “Order Blocks, Explained Properly” test, not to guess the next market move.Use the FXAbsolute AI Backtesting Lab to inspect calculations connected to “Order Blocks, Explained Properly” and the assumptions behind them.Reproduce any important “Order Blocks, Explained Properly” result and reserve unseen data before deciding that an apparent pattern is useful.
Continue your exploration of Order Blocks, Explained Properly with the beginner AI prompt guide, or inspect public calculations in the AI Backtesting Lab.