An ICT rejection block is a price zone framed by a swing's wick extreme and its most extreme opening or closing price. A bearish block sits between the highest body reference and the highest wick; a bullish block sits between the lowest wick and the lowest body reference. The formation can involve one candle or several. Its boundaries identify an area to examine when price returns. An entry requires its own timing, order and failure conditions.
The important detail is which candles supply those two boundaries. The candle with the longest wick may not supply the most extreme body. Drawing around just that candle can produce a different rectangle, midpoint and trade. The constructed examples below keep those choices visible, then compare an early boundary entry with a later confirmation entry under the same hypothetical cash budget.
Read the body reference as well as the wick
For any completed candle, the body extends between its opening and closing prices. Its top is the higher of those two values; its bottom is the lower. The high and low include the wicks. A candle's color describes the relationship between open and close, rather than deciding which boundary you should use.
For a selected group forming a swing high, find the highest high and the highest opening or closing price within that group. Those values frame the bearish rejection zone. For a selected group at a swing low, find the lowest low and the lowest opening or closing price. Those values frame the bullish zone.
The original ICT rejection-block lesson discusses multiple candles and body-reference retests that do not require a fresh wick extreme. It also presents direct boundary entries and a return-through-the-boundary alternative. These are different entry conditions. A later structure-confirmation rule adds another condition rather than changing the original rectangle's coordinates.
The broader Smart Money Concepts guide connects liquidity, structure and entry zones. For this lesson, start with the two prices defining the block and the time at which you could have marked them. A useful label should let another reader reproduce those coordinates without knowing the later outcome.
Draw a bearish block across two different candles
Suppose you have selected two completed candles at a swing high. Candle A opens at 202, reaches 212, trades down to 198 and closes at 206. Candle B opens at 205, reaches 210, trades down to 201 and closes at 208. These are invented prices for checking the drawing, not a historical trade.
Candle A supplies the highest wick, 212. Its body top is 206. Candle B has a lower wick high, 210, but a higher body top, 208. Across the selected group, the highest opening or closing price is therefore 208. The group-based bearish rejection block is 208–212.
| Reference | Candle A | Candle B | Selected group |
|---|---|---|---|
| Open / close | 202 / 206 | 205 / 208 | Highest body reference: 208 |
| High / low | 212 / 198 | 210 / 201 | Highest wick: 212 |
| Upper wick on that candle | 206–212 | 208–210 | Group rejection zone: 208–212 |
| Zone midpoint | 209 for A's upper wick | 209 for B's upper wick | 210 for the group zone |
If you draw only Candle A's upper wick, you obtain 206–212. That is a six-point band instead of the group's four-point band. Both individual upper wicks happen to have midpoint 209 in this example, while the combined selection has midpoint 210. A midpoint is calculated from the chosen boundaries; it does not identify those boundaries for you.
Record the selected candles and stop extending the group once the setup definition has been frozen. Otherwise, adding a later candle can shift the body boundary after you have already seen whether a particular entry would win. If your method deliberately updates the group, record each update and the orders affected by it.
A body-reference retest can stay below the old high
Continue the bearish example after price has moved away from the marked group. Imagine a later candle that opens at 202, reaches 210, trades down to 201 and closes at 207. It has traded above the body's 208 reference and then closed below it. Its high of 210 remains below the old wick high of 212.
That return has entered the 208–212 band without making a higher wick high. A study requiring a new break above 212 would reject it. A study examining a return through the 208 body reference could accept it as a candidate. The difference follows from the event being tested, rather than from one reader spotting a hidden candle.
Keep a sweep of a separately marked old high distinct from this later body-boundary retest. The liquidity-sweep guide uses a declared high or low and a reclaim condition. Its sequence can provide context for a rejection zone, but the two observations have different reference prices.
A wick on its own also cannot reveal the identities or intentions of the participants trading there. You can record that price reached 210 and returned below 208. Proving that a particular institution deliberately triggered stops would require additional evidence. The measurable event is enough to build a testable candidate.
Mirror the calculation for a bullish block
For the bullish example, select Candle A with open 98, high 102, low 88 and close 94. Select Candle B with open 95, high 99, low 90 and close 92. A supplies the lowest wick, 88. B supplies the lowest opening or closing price, 92. The group-based bullish zone is 88–92.
Using only A's lower wick would produce 88–94. The wider single-candle band has midpoint 91, whereas the group band has midpoint 90. Here too, the most extreme wick and body come from different candles. The same selection rule works despite the change in direction.
After price leaves the group, imagine a later return candle opening at 98, reaching 99, falling to 90 and closing at 93. It has crossed below the 92 body reference and returned above it. Its low remains above 88. This is the bullish counterpart of the earlier body-reference retest, rather than a fresh sweep below the group's wick low.
A return can also continue through 88. Save that outcome with the same care as the clean reaction. The rectangle identifies the area you selected; it does not compel price to reverse there or tell you how many retests it can survive.
Separate the formation, return and entry clocks
There are at least three useful timestamps: when the selected formation is complete, when price later returns, and when your chosen entry condition becomes available. Keeping these separate prevents a common replay error: entering at a price touched before the evidence needed for that entry existed.
For the bearish example, one hypothetical method places a sell limit at 208 after the formation and a move away, before the later return. Another waits for the returning candle to finish at 207 below the body boundary, then submits a sell stop at 206.5. That stop order becomes eligible only after the confirming close.
The sell limit can participate earlier and receives a different price if filled. The sell-stop variant asks price to resume downward after the returning candle has closed. It can remain unfilled if price never reaches 206.5 after submission. A move through 206.5 earlier in the same return candle does not fill an order submitted later.
The bullish mirror is a buy limit at 92 placed before the later return, or a buy stop at 93.5 submitted after the returning candle closes at 93. Use the same timing standards in both directions. A different rule for longs and shorts is possible, but it becomes another hypothesis to record and evaluate separately.
You can instead require a lower-timeframe structure event. The BOS, CHoCH and MSS guide explains those conventions. Name the swing, the required break or close and its completion time before comparing results. “Wait for confirmation” needs a specific observable event to become an executable rule.
Compare entry choices using the same cash budget
Hold the bearish protective stop at 213, one point above the selected wick high, and the target at 196. The one-point distance outside the zone is an invented example parameter. It is not a recommended buffer for a particular market. Assume one hypothetical unit is worth $2 per point, total round-trip costs are $2 per unit, and every entry and exit fills exactly as specified.
At the 208 boundary entry, the price risk is five points and the reward distance is twelve. Gross reward divided by risk is 2.4. The modeled loss per unit is five times $2 plus $2 in costs, or $12. A target fill would produce twelve times $2 minus $2, or $22 net per unit.
At the later 206.5 entry, the unchanged stop is 6.5 points away and the unchanged target is 10.5 points away. Gross reward divided by risk is about 1.62. Modeled loss per unit rises to $15, while net profit at the target falls to $19. Waiting for that particular event has changed the available distances. Whether its additional information compensates for those changes requires actual testing.
| Hypothetical short | Boundary limit | Later sell stop |
|---|---|---|
| Entry / stop / target | 208 / 213 / 196 | 206.5 / 213 / 196 |
| Risk / reward in points | 5 / 12 | 6.5 / 10.5 |
| Gross price reward-to-risk | 2.40 | About 1.62 |
| Modeled loss per unit, including costs | $12 | $15 |
| Whole units within a $90 budget | 7 | 6 |
| Total modeled loss | $84 | $90 |
| Net result if every unit reaches target | +$154 | +$114 |
The unit count is the cash budget divided by modeled loss per unit, rounded down. Seven boundary units use $84 of the $90 allowance; eight would require $96. Six later-entry units use $90. Spending the entire budget is not necessary for a valid calculation.
These are planned outcomes under invented assumptions, not executed trades, quoted contract specifications or a performance comparison. Slippage, gaps and partial fills can change them. For a real instrument, replace the point value and costs with its applicable specifications and execution assumptions.
The bullish mirror uses entries 92 or 93.5, stop 87 and target 104. The risk and reward distances match the bearish examples, so the same hypothetical unit counts and cash results follow. This symmetry checks the arithmetic; it does not predict equal real-world results for buying and selling.
Give the stop and the zone separate jobs
A protective stop controls the proposed position's exit. A zone-invalidating condition decides whether a chart idea remains eligible for future entries. They can use different prices or observation times. A completed-close rule for the zone does not postpone an already active protective stop.
Suppose the bearish plan has an active stop at 213 and a separate study rule that cancels the zone only after a completed close above its far edge. Price can reach the protective stop before that close exists. Record the position's exit under its execution assumptions, then evaluate the zone condition at its own timestamp.
If you replace the 213 stop with a closer lower-timeframe swing stop, recalculate the size and describe the changed invalidation premise. Keeping the original size while moving the stop is no longer the same $90-budget comparison. Moving the stop farther away after entry changes the planned loss as well.
A midpoint entry is another independent choice. The bearish group's midpoint is 210. A sell limit there will not fill on a return that reaches only 209. Calling that return a winning trade because price later falls to 196 counts a fill that never occurred. Preserve the unfilled record alongside the directional observation.
Compare rejection blocks with other marked zones
An order block selects a candle or candle group under a defined displacement and structure convention. The order-block guide compares body and full-range drawings. A rejection block instead focuses on the selected swing's wick-to-body band. Overlapping rectangles do not make their formation tests interchangeable.
A breaker block requires the relevant earlier order block and its specified failure, followed by the role-change sequence being studied. A rejection wick is not enough to establish that history. An FVG uses a defined non-overlap between the first and third candles of a three-candle formation. Its coordinates answer a different question from the highest body and wick at a swing.
These objects can coexist. A departure from a rejection area may leave a gap and break a smaller swing. Record the rejection band's origin, then record when the additional objects become available. Requiring every feature retrospectively can select attractive examples while concealing how few candidates were actually tradable in real time.
Test one additional condition at a time. For example, compare the same boundary method with and without a declared structure close. Keep the selected group, cost assumptions and target rule constant. Otherwise, a different result cannot tell you whether the added condition helped or whether you changed the entire trade.
Choose observation settings that you can reproduce
Record the instrument, data feed, formation timeframe and session settings with each zone. The same market on two feeds can show slightly different highs and closes. A narrow body-to-wick band makes those differences relevant to whether an order price was reached.
Also record how you identify a swing. A pivot definition requiring later candles becomes confirmed later than a method that marks a developing extreme immediately. If an indicator places its label back on the earlier candle, its visual location may precede the time when the label was knowable.
For a higher-timeframe formation with a lower-timeframe entry, save both clocks. A candle that currently has a long wick can extend before it closes. A completed formation rule must wait for that close; a developing-candle method needs its own update and cancellation policy.
Session, scheduled announcements and volatility can be useful test fields. Define any exclusions before reviewing the results. A long wick near an announcement is an observation about that price interval, rather than proof that the same order could have filled at the displayed level. Use appropriate data and execution assumptions for the conditions being studied.
Turn the drawing into a reviewable record
Start with a consecutive set of observation windows. Save the selected candles, both boundary prices, the formation time and the planned target before revealing the later return. Record whether the zone was reached, whether the trigger appeared and whether an eligible order actually filled.
Keep no-return windows, returns without confirmation, unfilled orders, stopped trades and unresolved sequences in the record. Each answers a different question. The fraction of marked zones that react is not automatically the win rate of a strategy with later entry, costs and a protective stop.
Separate the dates used to develop a rule from dates used to check it afterward. If you change the group selection or midpoint treatment during review, start a new version. Keeping the less attractive earlier version makes the comparison understandable.
The original rejection-block worksheet contains the two-candle boundary calculation and fields for the formation, return, order eligibility and outcome. It is an independently written practice sheet. The examples make no claim about a fixed win rate or the profitability of the broader method.

Compare HDP discounts
Lucid 30% · Tradeify 30% · TPT 50%.
See current offers ↗Eligible partner plans only.
Frequently asked questions
What is an ICT rejection block?
It is a selected swing's wick-to-body zone. A bearish group uses its highest wick and highest opening or closing price; a bullish group uses its lowest wick and lowest opening or closing price. The drawing and trade-entry rule are separate.
Can a rejection block contain several candles?
Yes. Freeze the selected group, then find its extreme wick and body reference. They can come from different candles. Using just the candle with the extreme wick can produce different boundaries and a different midpoint.
Does the rejection candle have to be bullish or bearish?
The boundary calculation uses prices rather than a compulsory candle color. For a bearish zone, find the highest open or close in the selected swing group. For a bullish zone, find the lowest open or close.
Must every retest make a new wick high or low?
No. In the bearish example, a return to 210 trades above the 208 body reference while remaining below the old high of 212. A model requiring a new wick-extreme sweep tests a different event.
How do I draw a bullish rejection block?
Select the completed swing-low group and mark its lowest wick and lowest open or close. With the two invented candles in this lesson, those prices are 88 and 92. Record the group's completion time before studying a later return.
Is a rejection block the same as a pin bar?
A pin bar describes a candle shape under the convention being used. A rejection-block study adds a selected swing, explicit zone coordinates and context. The presence of a long wick alone does not specify an executable entry.
Should I enter at the boundary or wait for confirmation?
They are separate variants to compare. The boundary method can place an order before the later return. A confirmation method waits for a defined event, then submits its order. Compare eligible fills, stop distances, costs and unfilled candidates.
Where does the protective stop belong?
The stop follows the model's invalidation premise. The worked bearish comparison uses 213 above the 212 wick high; its bullish mirror uses 87 below 88. Those invented buffers are not instrument-specific recommendations, and exact fills are assumptions.
Is the midpoint always the best entry?
No universal best entry follows from the midpoint. The bearish group midpoint is 210. An order there misses a return that reaches only 209. A deeper entry changes the eligible trade set as well as the planned reward-to-risk.
How is a rejection block different from an order block?
A rejection block uses a selected swing's wick and body references. An order block uses its declared origin-candle, displacement and structure rules, with a chosen body or full-range boundary convention. Compare those rules and coordinates directly.
Does a failed rejection block automatically become a breaker?
No. A breaker needs the relevant earlier order block and its specified failure and role-change sequence. Price crossing a rejection zone alone does not establish that history or identify a new trade.
Which timeframe is best for rejection blocks?
This lesson establishes no universally best timeframe. Record the formation and entry intervals, session and swing-confirmation delay, then evaluate one defined method. A larger rectangle can require a larger stop distance and a different position size.
Do I need a rejection-block indicator?
You can calculate the boundaries manually. An indicator adds its own group-selection, swing-confirmation and invalidation rules. Check when its signals become available and whether historical labels are drawn earlier than that time.
What is the win rate of the rejection-block strategy?
A fixed win rate cannot be inferred from a label or selected chart examples. Define formation, eligibility, fills, stops, targets, costs and the sample period, then evaluate the complete record. The worksheet provides practice fields rather than a performance claim.
