Where to Take Profit on ICT and CRT Setups

Most traders can tell you why they entered. Ask where they're getting out and you get a shrug, a round number, or "I'll see how it looks."

That's the leak. The entry decides whether you're in a good trade; the exit decides what that trade is worth. And unlike the entry, the exit gets made while you're already in the position, with money moving on the screen — which is exactly the wrong condition for making a decision you hadn't planned.

This is the other half of where to place your stop loss. Same principle, mirrored: the chart tells you where you're going, before you enter, and you write it down.

The one question your target answers

A stop answers "at what price is this idea wrong?" A target answers a different question, and it isn't "how much do I want to make":

Where is the next pool of resting orders that price is likely to reach for?

That's it. Smart money setups are built on a single premise: price moves between areas where orders are resting, because that's where the fills are. If you take an entry on that premise, your exit has to obey the same logic. A target picked because it's "a nice 3R" is a target with no theory behind it — you've used a liquidity model to get in and a wish to get out.

So before you enter, you should be able to point at a specific level and say that's what this move is going for. If you can't find one, that's not a trade with an unclear target. That's not a trade.

Targets by setup

Liquidity sweep and CRT

The setup: price runs an old high or low, takes the liquidity resting there, and closes back inside the range. You're in on the reversal.

The target is the opposite side of the range — the liquidity on the other end that hasn't been taken yet.

This is the cleanest target in smart money trading, because the whole CRT model is a statement about exactly this: the range gets built, one side gets swept for fills, and expansion runs toward the other side. The sweep told you which direction. The opposite edge tells you where the move has a reason to stop.

Two refinements worth having as written rules:

• Equilibrium is the interim level. The 50% of the range is where a large share of moves stall or retrace. If you're taking anything off early, that's the level with a reason behind it — not a random R multiple. See premium and discount for why the midpoint matters structurally rather than as a Fibonacci habit.
• Old highs and lows in the way are speed bumps, not the destination. If there's an untouched short-term high between your entry and the range edge, expect a reaction there. Whether you exit there or hold through it should be a rule you've tested, not a decision you make while watching it happen.

Order block mitigation

The setup: price returns to an unmitigated order block and you enter on the reaction.

The target is the liquidity the original move was heading for — usually the high or low that the displacement out of that block was going to take, and often didn't finish taking.

An order block exists because a move originated there. That move had a destination. When price comes back and mitigates the zone, the sensible expectation is that it resumes toward the destination that was already in progress. So the target isn't derived from your entry at all — it's inherited from the move that created the zone in the first place.

This is why marking the zone properly matters for the exit and not just the entry. If you don't know what the displacement was reaching for, you have an entry with no destination attached.

Fair value gap entry

The setup: price retraces into an unfilled imbalance and continues.

The target is the extreme of the leg that created the gap, and beyond it, whatever liquidity sits just past that extreme.

A fair value gap is the fingerprint of an aggressive move. That move stopped somewhere, and that stopping point is usually a swing high or low with orders parked beyond it. Price returning to the gap and continuing is the same participant finishing the job. Order blocks vs fair value gaps covers how the two zones relate — worth reading if you're targeting one from the other.

A trap to name: the next fair value gap in the opposite direction is a real obstacle. If there's an unfilled imbalance sitting between you and your target, price frequently rebalances there first. That's not a reason to skip the trade; it is a reason to know it's there before you're staring at a stalled position.

Breaker block

The setup: a level failed, structure shifted, and the zone now works in the other direction.

The target is the liquidity beyond the failed level — the stops of everyone who was positioned the old way.

A breaker is created by a group of traders being wrong in a visible place. Their stops are the fuel for the next move, and they sit just past the level that broke. That's a specific, findable price, which is exactly what you want in a target.

The three kinds of target, and when each is right

Target type	Where it comes from	Use it when	How it fails
Structural liquidity	An old high/low, the opposite range edge, an obvious equal-highs cluster	Default. It's the only target with a mechanism behind it.	Sometimes far away, producing a great R you never get paid
Imbalance / equilibrium	Unfilled FVG, 50% of the range, the origin of a displacement	As an interim level, or when structural liquidity is unrealistically distant	Fills partially and reverses; easy to over-use as an excuse to exit early
Fixed R multiple	Your own risk unit — exit at 2R, 3R	When you're testing and need every trade measured identically, or when structure genuinely offers nothing	Has no market logic at all. Price doesn't know where your stop was

Most traders should be running the first as the default, the second as the interim, and the third only inside a test. If you're using fixed R live because it's simpler, be honest that you've swapped a model-based exit for an arbitrary one — and that a 2R target on a setup whose real destination was 5R away is a systematic leak that no win rate fixes.

Partial exits: what scaling out actually does

Everyone scales out and almost nobody has thought about what it does to the numbers. Here it is plainly:

Taking partial profits lowers your variance and lowers your expectancy. Both. It is not free. You're selling part of your right tail — the small number of trades that run far — in exchange for a smoother equity curve and less discomfort.

That trade can be worth making. It is worth making when:

• You're new to the setup and holding a full position to a distant target makes you break rules you'd otherwise follow. A slightly worse expectancy you can actually execute beats a better one you can't.
• Your target is genuinely uncertain — for example the structural level is far and there's an obvious interim obstacle in the way.
• You're trading a size where a full stop-out matters to you personally rather than just statistically.

It is not worth making when it's a habit you've never measured, and that's the usual case. A common pattern: exit half at 1R, move the stop to entry, then watch the remainder get stopped at break-even on the retrace that was always going to happen. That combination systematically converts winners into fractions of winners and keeps every full loser. Run the numbers on it before assuming it's prudent — the risk/reward calculator will show you what break-even win rate a given R actually demands, and halving your average R raises that bar a lot more than people expect.

If you do scale out, make it a rule, not a reflex: which level, what fraction, and what the remainder's stop does. Written down, applied identically, testable.

The minimum-R filter — a decision made before entry

Once you have an honest stop and an honest target, you have an R multiple you didn't choose. Now use it as a filter.

Set a minimum. If the structural target is 1.2R away from a properly-placed stop, that's a setup your strategy shouldn't take, no matter how clean it looks. And notice what you must not do about it: you may not move the stop closer to manufacture the ratio. That converts a trade with real invalidation into a coin flip, which is covered at length in the stop loss guide.

Run position size from the honest stop and check the ratio from the honest target. Two numbers, both known before you click, and together they answer whether the trade exists. Most traders compute neither and then wonder why a good-looking setup lost money.

When price stalls before the target

This is the case nobody writes rules for, so it gets improvised — and improvised exits are where discipline goes to die. Decide these in advance:

Structure-based exit. If price shifts structure against you before reaching the target — a lower high forming on a long, a break of the swing that was carrying the move — the premise has changed. That's a legitimate exit, and it's not the same as bailing because the position went red for ten minutes.

Time-based exit. Many intraday smart-money setups are session ideas. If the move hasn't materialised by the time the session that motivated it is over, you're now holding a trade for a reason that no longer applies. A rule like "flat by the end of the session I entered in" is crude but it's better than holding indefinitely because you don't want to book a small loss.

Reaction at an interim level. If price reaches equilibrium or an intermediate high and reacts hard — a strong rejection, an imbalance forming against you — that's information. Whether you act on it must be a rule you tested, because "it looked weak" is available on every trade you're bored of.

Doing nothing is also a decision. If none of your written exit conditions have triggered, hold. The most expensive exits are the ones motivated by the position's P&L rather than by anything on the chart.

How to find out which exit rule is actually best

Here's the part that turns all of this from opinion into your own data, and it's one extra column in your journal.

For every trade, log the furthest price went in your favour before it hit your stop or your target. Do it in R: "this trade went +2.8R before it came back and stopped me at −1R."

Collect thirty or forty of those and you can answer questions no article can answer for you:

• Are your targets too far? If most trades reach 2R and reverse before your 4R target, you're systematically giving back moves that were paid to you.
• Are your targets too close? If a large share of your winners kept running well beyond your exit, your scaling-out habit is costing more than it saves.
• Which setup type behaves differently? Sweeps and mitigations don't have the same shape. Segment before concluding.

This is exactly the review loop in how to actually learn a trading strategy, applied to the exit rather than the entry — and the exit is where it's almost never applied. How to journal your trades covers the rest of the fields; add this one and your exits stop being a matter of taste.

One caution while you gather that data: change one exit rule at a time. If you move the target and start taking partials in the same batch, you'll have no idea which did what.

The mistakes worth naming

• No target before entry. If you can't name the destination, the setup isn't complete. Entry-only trading is how people end up managing positions by feeling.
• Round numbers as targets. 20,000 on an index and 1.1000 on EURUSD are where everyone's take-profits sit — which is a liquidity pool, not a destination. Being just in front of one is often smarter than being at it.
• Moving the target further away while in profit. The mirror of widening a stop, and the same error: rewriting the plan to match the outcome you want.
• Cutting winners at the first sign of a pullback. Retracement into the entry zone is normal on almost every smart-money setup. If you exit on it, you'll never see the expansion your entire model is about.
• Using one target rule across every setup. A sweep on a defined range and a mitigation with a far-away objective are different trades. One fixed rule flatters one and ruins the other.
• Judging exits by the last trade. Every exit rule feels wrong immediately afterwards — you either left money on the table or gave some back. Judge over a sample or you'll rewrite your rules weekly forever.

Where this fits

Entry, invalidation, destination. Those are the three parts of a trade, and most people learn them in the order entry → invalidation → nothing. Getting the third one right doesn't require a new strategy — it requires marking the liquidity you were already using to justify the entry, and treating it as a level you'll act on rather than a story you told yourself.

If the earlier steps are still fuzzy — which liquidity matters, what a valid displacement looks like, how to mark a zone — start with ICT concepts explained for beginners, then the CRT mitigation strategy for the specific sequence this exit logic was written for.

And when you'd rather learn the whole thing as one system — the sweep, the structure shift, the entry, exactly where risk sits and exactly where the trade is going — that's what the CRTLAB course teaches, end to end, one payment, lifetime access. Exits are the part free content skips most, because they're the part that only makes sense once the model underneath them is complete.

FAQ

Where should I take profit on a liquidity sweep? At the opposite side of the range — the liquidity that hasn't been taken yet. The sweep tells you the direction; the untouched side tells you what the move is reaching for. Equilibrium at the 50% of the range is the natural interim level if you're taking anything off early.

Should I take partial profits? Only as a written rule, and knowing what it costs. Scaling out reduces variance and reduces expectancy — you're selling part of your right tail for a smoother curve. It's a fair trade when a distant target makes you break rules you'd otherwise follow; it's a leak when it's an unmeasured habit, especially combined with moving the stop to break-even.

What's a good risk-reward ratio for ICT setups? The one the chart gives you, filtered by a minimum you set in advance. Place the stop where invalidation actually is, find the structural target, and see what ratio falls out. If it's below your minimum, skip the trade — don't tighten the stop to manufacture a better number.

Should I use a fixed R target instead of structure? Only inside a test, where measuring every trade identically has value, or when structure genuinely offers nothing. Live, a fixed R target has no market logic behind it — price has no idea where your stop was — and it systematically caps trades whose real destination was much further away.

Where do I take profit on an order block entry? At the liquidity the original displacement out of that block was heading for — typically the high or low that move was going to take. The target is inherited from the move that created the zone, not derived from your entry.

What do I do if price stalls before hitting my target? Follow a rule you wrote before entering: a structure shift against you, the end of the session that motivated the trade, or a strong rejection at an interim level. If none of those has happened, hold. Exiting because the position's P&L is uncomfortable is not an exit rule.

How do I know if my take profit is too close or too far? Log how far each trade went in your favour before it resolved, measured in R. After thirty or forty trades the answer is in your own data: winners that ran far beyond your exit mean you're leaving money behind, and trades that repeatedly stall just short of your target mean it's too ambitious. Segment by setup type before concluding.

Should I move my stop to break-even when I take partials? Be careful — that combination is one of the most common ways traders convert winners into break-evens. Most smart-money setups retrace into the entry zone before expanding, so a mechanical break-even removes you from exactly the trades that were about to work. Move the stop when structure justifies it, not when the first partial fills.