How to Cast an I Ching Hexagram: The Three-Coin Method, Step by Step

The three coins land on the table, you add the values, and draw one line. Do that six times and a hexagram stands in front of you - six lines stacked from the bottom up, each one either broken or solid, some of them marked to change. That small procedure is the modern doorway into the I Ching, the Book of Changes, a Chinese classic whose core, the Zhou Yi, is built from 64 hexagrams and the texts attached to them. The casting takes five minutes. Learning to do it cleanly, so the figure on paper matches what the coins actually said, takes one careful read-through and a bit of practice.

The method below is mechanical on purpose. Toss, add, draw, repeat. Interpretation comes after the figure exists, and it works better when the figure was recorded without wishful editing along the way. Readers who want to try the cast in a digital form can use the I Ching oracle on this site; readers working with physical coins will find every step here, including a worked example you can check your own casts against.

What the I Ching is, briefly

A hexagram is a figure of six horizontal lines. Each line is either yang, drawn as one unbroken stroke, or yin, drawn as two short strokes with a gap in the middle. Every hexagram is two three-line figures, called trigrams, stacked: a lower trigram (lines 1 to 3) and an upper trigram (lines 4 to 6). There are eight possible trigrams, so there are 64 possible hexagrams.

Each hexagram carries a short text called the Judgment, and each of its six lines carries its own line text. Tradition credits the Judgments to King Wen of Zhou and the line texts to his son, the Duke of Zhou; that is the traditional attribution, passed down with the text itself. The language of these texts is terse, image-heavy poetry, with roots in early Chinese statecraft: river crossings, waiting, armies, marriages, harvests, audiences with rulers. Later centuries added commentarial wings, but the casting method works on the core: 64 figures, their Judgments, and their line texts.

The cast answers nothing by itself. It selects which of the 64 figures, and possibly which of its lines, you will sit with. The rest of this guide covers the selection procedure properly, because a sloppy cast produces a random hexagram plus recording errors, which helps nobody.

What you need

• Three coins of the same type, with a clear heads side and tails side
• Paper and a pen
• A flat surface the coins can land on without rolling off
• The question, written down (more on wording below)
• A translation or reference for the hexagram texts, or the site oracle linked above

That is the whole kit. Some practitioners keep dedicated coins; dedicated coins are a habit some people like, not a requirement of the method. The pen matters more than the coins. Every toss gets written down as a number the moment it lands, because memory edits sequences within minutes.

The coin values

Assign a number to each side before you begin:

• Heads = 3
• Tails = 2

Some schools flip this assignment. The totals method works identically either way, so pick the convention above and keep it for the whole cast. Toss all three coins at once and add the three values. The sum is always 6, 7, 8, or 9, and each sum names one kind of line:

Old and young describe the line's state in the tradition's picture of change: a young line is settled, an old line has reached its extreme and flips into its opposite. That flip is what generates a second hexagram later. For now, the only habit to build is marking: every 6 gets an x, every 9 gets a circle, the moment the line is drawn. Do not trust yourself to remember which lines changed after six tosses. Nobody remembers reliably, which is why the marks exist.

Building the six lines

One toss makes one line. Six tosses make the hexagram. The first toss is line 1, at the base. The second toss sits on top of it as line 2. Continue upward until the sixth toss becomes line 6, the top.

A practical rhythm helps:

1. Shake the three coins in cupped hands, thinking of the question as written.
2. Let them fall on the surface.
3. Read the sides without rearranging them, add the values, say the total quietly.
4. Write the total down, then draw the line on the stack, with its mark if it changes.
5. Pick the coins up and repeat, five more times.

Work bottom to top on the paper. Many beginners draw their first line at the top of a fresh page and build downward, then copy the figure wrong when they flip it. Save yourself the error: leave space, start low.

After six tosses you hold a list of six totals and a drawn hexagram. The list of totals is the cast's permanent record. Given the totals, anyone can rebuild the hexagram, identify the changing lines, and check the reading. Given only a drawing made in a hurry, nobody can.

Finding the hexagram: trigrams

With six lines drawn, split them into two trigrams. Lines 1 to 3 form the lower trigram, lines 4 to 6 form the upper. Read each trigram from its own bottom line up, and match the pattern:

The pair of trigrams names the hexagram, and the hexagram's position in the sequence of 64 gives its number in the received King Wen order, which is the numbering translations use. Fire under Lake, Lake over Mountain, and so on; the upper trigram is spoken first in the traditional description, as in Thunder over Fire. At this stage, a reference table or the oracle page does the lookup for you. What matters for a clean cast is that your two trigram patterns are read in the right direction. Bottom-to-top, every time.

A worked example, end to end

Suppose the question has been written down and six tosses produce these totals, in order, first toss first:

Lines bottom to top: yang, yin, old yang, yang, yin, old yin.

Lower trigram, lines 1 to 3: yang, yin, yang. Against the table, that is Li, Fire. Upper trigram, lines 4 to 6: yang, yin, yin. That is Zhen, Thunder. Thunder over Fire is hexagram 55, Feng, Abundance, in the King Wen order.

Two lines change here: line 3 (the 9, old yang) and line 6 (the 6, old yin). Flip each changing line into its opposite and leave the stable lines alone. Line 3 drops from yang to yin. Line 6 rises from yin to yang. The transformed hexagram, bottom to top, reads: yang, yin, yin, yang, yin, yang.

Split it again. Lower trigram yang, yin, yin: Zhen, Thunder. Upper trigram yang, yin, yang: Li, Fire. Fire over Thunder is hexagram 21, Shi He, Biting Through.

So this cast reads in full: primary hexagram 55, changing lines 3 and 6, moving toward hexagram 21. That three-part summary is exactly what goes at the top of a page of notes, above any interpretation. If a cast cannot be summarized that way, the recording went wrong somewhere, and the fix is to rebuild from the written totals, not to recast.

Checking the cast before you read anything

Pause between casting and reading. A two-minute check at this point saves the most common failure in home casts, which is not a bad question or a poor translation but a figure copied wrong between the table and the page. Run the check in the same order every time, and it becomes part of the rhythm.

First, read the six totals aloud from the written list, lowest first. Each number should be 6, 7, 8, or 9. If a 5 or a 10 appears, a coin was misread or two tosses were merged in the notes; the totals list no longer matches the coins, and the cleanest fix is to note the error and cast again rather than guess which toss went wrong.

Second, rebuild the drawn stack from the totals without looking at the drawing. Totals of 7 and 9 become solid lines; 8 and 6 become broken lines. Compare the rebuilt stack with the one on the page. They should match line for line. Where they differ, trust the totals, because the totals were written at the moment each toss landed, while the drawing was made seconds later each time.

Third, confirm the marks. Every 6 carries its x mark, every 9 its circle, and no other line carries anything. A missing mark will not change the primary hexagram, but it will quietly change the reading, because an unmarked changing line never gets flipped and its line text never gets read in its proper place.

Fourth, split the stack into trigrams again and name them out loud, lower then upper, each read from its own base upward. Many lookup errors trace back to reading the upper trigram downward from the top of the page, which mirrors the pattern and can produce a different trigram with a similar feel. Reading patterns in their built direction, lowest line first, is the whole discipline of the method in miniature.

Only after those four checks should the book or the oracle page open. This pause has a second use: it lets the question settle. You wrote it down before casting; read it once more now, exactly as written, and let the reading address that sentence rather than the looser concern that prompted it.

Changing lines and the second hexagram

Any line cast as 6 or 9 changes, and when at least one line changes, the cast produces a second, transformed hexagram. The primary hexagram describes the situation as it stands in the frame of the reading. The transformed hexagram is read afterward, as where the situation is read as moving. It is not a replacement and it is not a better answer; it comes second in the classical reading order and carries less weight than the primary figure.

A few structural points save confusion:

• Only 6s and 9s flip. Young lines (7 and 8) stay exactly as drawn in both hexagrams.
• One changing line is the most common case. Several can change at once, as in the worked example, and occasionally all six change, which flips every line.
• If no line changes, there is no transformed hexagram. The cast is a single figure, and nothing was done wrong. Some sitters feel short-changed by a static cast; the tradition treats it plainly, as a situation holding its shape for now.

How the reading is put together

The classical reading order has three steps when lines change, and one when they do not:

1. The primary hexagram and its Judgment. Read the hexagram name, its image, and the Judgment text first. This sets the frame for everything after it.
2. The changing lines. Read the line text of each changing line, in order from the bottom. These speak within the frame, and in practice they often carry the most specific language in the reading.
3. The transformed hexagram's Judgment. Read it last, as the direction the situation is read as moving toward.

With no changing lines, read the primary Judgment and stop there.

Two habits keep this order clean. First, read all three layers before interpreting any of them; a line text quoted against its hexagram reads differently than a line text floating free. Second, write the interpretation down in your own words the same day. Casts blur in memory, and a dated sentence about what you took from the reading is the only version you can learn from later. Readers interested in how another Chinese system organizes a reading around set structures can compare the pillar logic in the BaZi mastery and destiny blueprint article; the surface differs completely, but the habit of reading parts against a whole is shared.

Coins compared with yarrow stalks

The older classical method uses yarrow stalks: 49 stalks, set through a longer series of divisions and countings, three rounds of manipulation per line. It takes considerably longer, often twenty minutes or more for a full cast, where coins take a few minutes. Both methods produce the same four line outcomes - 6, 7, 8, 9 - and both reach the same 64 hexagrams.

The odds differ between the two methods. With three coins the arithmetic is plain: there are eight equally likely outcomes. A total of 6 comes only from three tails (1 way out of 8) and a total of 9 only from three heads (1 out of 8), while 7 and 8 each come from three arrangements (3 out of 8 each). Changing lines are therefore less frequent than stable lines with coins, in a fixed, knowable ratio. The yarrow procedure skews the same four outcomes differently; the stalk division process does not weight the four results the way coin combinations do. Practitioners argue about which probability profile the oldest texts assumed. For a beginner that argument can wait; what matters is choosing one method and learning its feel.

Coins have two practical advantages: speed, and a cast that is easy to audit afterward from the written totals. Yarrow has one: the slow handling forces a still, deliberate pace that many readers value. Neither advantage is mystical, and the texts read the same either way.

Framing the question

Customary practice in the tradition asks for one question, written down, in a calm setting. That is a description of how practitioners usually work, and each part of it does practical work:

• One question. Two questions in one cast produce a reading that fits neither, because every text gets stretched between two subjects. Split compound questions before casting, not after.
• Written down. Writing the question forces it into a finished sentence. Vague unease rarely survives being phrased, and the written wording is what the reading gets measured against later.
• A calm setting. In the tradition, a settled room and an unhurried pace are customary. Practically, rushing correlates with recording errors: totals misread, lines drawn in the wrong order, marks forgotten.

Questions framed around your own conduct and timing - what to pay attention to, what approach fits this situation - take to the method better than demands for a verdict on someone else. The texts speak in images of conduct far more often than they name external events. Similar care with question framing shows up in other card and chart traditions covered on this site, such as the three-card spread reading guide and the cartomancy study guide, where the spread or chart also works best against one finished sentence of a question.

Limits, stated plainly

The method is a structured reflection tool in modern use. The book at its center is poetry and statecraft from early China, and readers treat the cast as a prompt: a selected set of images and counsel to think with, weighed by the reader, in the reader's own situation. Tradition frames the cast as a response to the question; no part of the procedure tests whether that response predicts anything, and this guide makes no such claim. A cast is not a verdict, not evidence about another person's intentions, and not a substitute for the ordinary work of checking facts, prices, contracts, and advice from qualified people.

There are also mechanical limits. The procedure selects a hexagram; it cannot flag a badly worded question, a wrong recorded total, or a translation that flattens an image into mush. Translations vary, sometimes a lot, and the same line text can read gently in one rendering and sternly in another. Working with one translation until its voice is familiar beats sampling five. And probability deserves one last mention: with 64 hexagrams and changing lines in play, a cast that feels strikingly apt will arrive now and then by chance alone. Readers in the tradition have always known the arithmetic; the practice continues anyway, as reflection, not as measurement.

Keep the kit simple. Three coins, a written question, six totals in a column, marks beside the changing lines, the primary figure read before the transformed one. Do that cleanly a few dozen times and the procedure disappears into the background, which is where it belongs - leaving the old, strange, image-heavy text to do the talking.