What Is CRT (Candle Range Theory)? A Plain-English Guide for Traders

If you spend any time around smart-money or ICT-style trading, you'll run into Candle Range Theory (CRT) fast. The name sounds technical, but the core idea is simple — and once it clicks, you start seeing it on almost every chart.

This guide covers the concept in plain English, then goes past the definition into the parts that actually decide whether you can trade it: which side of the range gets swept, which candle you should be using, and what tells you the idea is wrong. If you want the condensed, definition-first version instead, that lives in the CRT reference entry.

What is Candle Range Theory?

Candle Range Theory is a way of reading price that uses the range of a single candle — its high to its low — as a map for what price is likely to do next.

Instead of treating a candle as just one bar on a chart, CRT treats that candle's range as a zone: a high, a low, and a midpoint. Price then interacts with that zone in a repeatable way — reaching for one edge to grab liquidity before expanding toward the other.

In one sentence: one candle sets the playing field, and the candles that follow tend to sweep one side of it, then run to the other.

The three phases, in the order they happen

Picture a single higher-timeframe candle. A CRT setup moves through three phases, and the order is fixed.

1. The range. A candle forms with a defined high and low. That range becomes your reference — the area price is working within. The midpoint (the 50% level) acts as equilibrium: fair value sitting between the two extremes.

2. The sweep. Price pushes beyond one edge of the range — above the high or below the low — taking out the orders resting there. This is the part that traps people: it looks like a breakout, but it's a liquidity grab, not a genuine directional move.

3. The expansion. After the sweep, price reverses and expands toward the opposite side of the range. The edge that got swept becomes the origin of the move; the untouched edge becomes the target.

Build a range, sweep one side, expand to the other. That rhythm is the heartbeat of CRT, and it's closely related to the Accumulation → Manipulation → Distribution (AMD) idea, sometimes called the Power of Three.

The sequence being fixed matters more than it sounds. A range with no sweep isn't a CRT setup. A sweep with no expansion is an incomplete one. Most of the bad trades taken in this model's name are one of those two things, taken as though they were the whole thing.

Which side gets swept?

This is the question the three-phase diagram never answers, and it's the one that decides whether you're on the right side of the move.

CRT tells you what shape to expect. It does not, on its own, tell you which direction. A range has two edges, and if you're flipping a coin on which one goes first, the cleanest model in the world won't save you.

Direction has to come from outside the range — from the higher timeframe. The usual reading is to ask where price is being drawn to: which side has the more obvious pool of resting orders, and where has price been failing to reach. An old high that's been approached three times and never taken is a magnet. A low that's already been swept twice has less business left.

The practical version: establish your directional read first, then use the range for the entry. If you find yourself picking direction from the range itself, you've inverted the process — and a range in isolation is a coin flip with extra steps. Market structure and liquidity are the two things that supply that read.

Which candle sets the range?

CRT is timeframe-agnostic — the logic is identical on a weekly candle and a 5-minute one. That's a genuine strength, and also the most common source of confusion, because "it works on any timeframe" gets misread as "it works on all of them at once."

In practice traders pair two timeframes: a higher one to define the range, and a lower one to time the entry inside it. The higher-timeframe candle is the playing field; the lower timeframe is where you watch the sweep and the shift happen with enough resolution to act on.

The pairing itself is a matter of preference rather than rule, but the principle isn't: the range candle should be high enough that its high and low are levels other people can see, and the entry timeframe should be low enough that the sweep is a readable event rather than a single wick. Two timeframes that are too close together give you a range and an entry that are effectively the same chart, which defeats the point.

Whichever pairing you choose, be consistent about it while you're testing. Switching the range timeframe mid-sample is the fastest way to produce results that measure nothing.

What invalidates a CRT setup

A model that can't be wrong can't be tested, so this is the section worth internalising.

The premise is: price sweeps one edge, then expands toward the other. That gives you two clean failure conditions.

The sweep doesn't reject. Price pushes beyond the edge and simply keeps going, accepting outside the range rather than snapping back. That's the scenario where what you called a liquidity grab was an actual expansion, and you were positioned against it. Trading the sweep as a reversal before price shows any rejection is the single most expensive habit in this model.

The expansion never comes. Price sweeps, rejects, and then does nothing — drifting sideways inside the range instead of running to the other edge. The idea isn't wrong exactly, it's just not paying, and a setup that needs to be right eventually is not a setup.

Both are worth knowing before entry, not after. If you can't state in advance what price has to do for you to be out, you don't have a trade — you have a hope with a level attached.

The ways CRT gets traded backwards

• Trading the sweep as a breakout. The move beyond the edge is the part CRT expects to fail. Entering on it is entering on the side the model exists to describe.
• Marking every candle's range. If twelve ranges are on your chart, none of them are meaningful. One range, from a candle that matters, in the direction you've already established.
• Taking direction from the range. Covered above, and worth repeating because it's the most common one.
• Treating the midpoint as a signal. Equilibrium is useful context for whether you're buying at a discount or a premium. It isn't an entry trigger by itself.
• Judging it on ten trades. CRT produces losing runs like everything else. A sample of ten tells you about variance, not about the setup.

How to build the recognition

Reading about CRT and trading it live are two very different things, and the gap between them is screen time — seeing the setup form in real time, enough times to trust it under pressure.

Explanation doesn't close that gap. Repetition does. The measurable way to get it is replay: step through historical price one candle at a time with the future hidden, mark the range, wait for the sweep, and log what the expansion actually did. Do that across 50–100 occurrences and you have a real win rate and average R for the setup instead of an impression of one — the step-by-step process is in How to Backtest the CRT Strategy, and the mistakes that quietly ruin a backtest are in this one.

You can practise CRT on real market history in CRTLAB for free — the CRT strategy backtester lets you pick a market, replay it bar by bar, and mark the range, the sweep and the expansion yourself, with every trade logged and scored as you go. It's the fastest way to turn "I understand CRT" into "I can trade CRT."

The bottom line

Candle Range Theory comes down to one repeatable idea: a candle's range sets the stage, price sweeps one side to grab liquidity, then expands toward the other. The model gives you the shape — direction, timeframe pairing and invalidation are what you bring to it. Get those three right, put in the screen time to trust the read, and you have a structured, testable way to approach the market rather than another pattern to hunt for.

FAQ

What does CRT stand for in trading?

CRT stands for Candle Range Theory. It's a way of reading price that uses a single candle's range — its high, its low and its 50% midpoint — as the reference zone for what price does next.

What are the three phases of CRT?

The range, the sweep and the expansion. A candle's high and low define the range, price then sweeps one edge to take the liquidity resting beyond it, and price then expands toward the opposite edge. The order is fixed: a range with no sweep isn't a CRT setup, and a sweep with no expansion is an incomplete one.

Is a CRT sweep the same as a breakout?

Close to the opposite. A sweep looks like a breakout but is expected to fail — its purpose is to take out the stop losses and breakout entries resting beyond the range edge before price moves the other way. Trading the sweep as a breakout is the most common way the model gets traded backwards.

Which timeframe is best for CRT?

Any, as long as you're consistent. The logic is identical on a weekly candle and a 5-minute one. Most traders pair a higher timeframe to define the range with a lower one to time the entry inside it, and keep that pairing fixed while they're testing.

How do you know which side of the range will be swept?

Not from the range itself — that's the mistake. Direction comes from the higher-timeframe context: where price is being drawn to, and which side holds the more obvious pool of untouched orders. CRT gives you the shape of the move; the directional read has to come first.

Is CRT the same as ICT?

No. CRT is a single focused model; ICT is a broad framework containing many concepts. They rest on the same liquidity premise, and CRT's three phases map closely onto ICT's Accumulation → Manipulation → Distribution, but CRT is one idea and ICT is a whole body of work.

How do you backtest CRT?

Replay historical price one candle at a time with the future hidden, mark the range, wait for the sweep, and log the outcome of the expansion by fixed rules. After 50–100 setups you have a measured win rate and average R rather than an impression of one.