Every tarot spread you have ever used was invented by someone. The Celtic Cross was popularised in print in the early twentieth century, the three card past present future layout is a convention rather than a law, and the elaborate ten card designs circulating online were mostly written by readers who wanted something their existing spreads did not provide. There is no authority issuing spreads. There is only whether a layout produces useful readings.
That said, most homemade spreads are bad, and they are bad in predictable ways. They ask overlapping questions, they include positions the reader cannot interpret, and they are far too large. A twelve position spread designed on a Sunday afternoon usually gets used once.
This guide gives you a five step method for designing a spread that survives repeated use, plus the position writing rules that separate a working layout from a decorative one.
Step One: Define the Job
Before you draw a single box on paper, finish this sentence: "This spread is for someone who wants to know..."
Be narrow. "A spread for love" is not a job, it is a category. "A spread for deciding whether to have a difficult conversation with a friend" is a job. Narrow spreads produce sharp readings, because each position can be tailored to a real situation instead of hedging across every possibility.
Ask yourself three follow up questions:
Who uses this, and how often? A weekly personal spread can be idiosyncratic. A spread you will use for other people needs positions that are explainable in one sentence.
What does the existing repertoire fail to do? If a three card spread already answers the question well, you do not need a new layout. Custom spreads are for gaps, not for novelty.
What should the reader walk away with? Understanding, a decision, an action, or a reframe. Each of those calls for a different final position.
Step Two: Choose the Card Count Honestly
The single most common mistake in custom spread design is size. More cards feels more thorough and reads as less clear, because each additional card dilutes the attention available for the others.
| Card count | Suits | Typical reading time |
|---|---|---|
| 1 to 2 | Daily focus, quick clarification | Under 5 minutes |
| 3 to 4 | Most situations, decisions, check ins | 10 to 15 minutes |
| 5 to 7 | Complex situations with several factors | 20 to 30 minutes |
| 8 or more | Deep dives, annual reviews, formal readings | 30 minutes or more |
Design for the smallest count that does the job. If your first draft has nine positions, look for the three that overlap and cut two of them. A well built five card spread beats a bloated ten card spread almost every time, and it gets used far more often.
The exception is a deliberately comprehensive layout like the Celtic Cross, where the size is the point. Even then, notice that each of its ten positions asks something genuinely distinct.
Step Three: Write Positions That Can Be Answered
A position is a question the card answers. If you cannot phrase the position as a clear question, the card will have nothing to respond to and you will end up reading a generic meaning into a meaningless slot.
Four rules for writing positions.
Rule one, one question per position. "What is blocking me and how do I fix it" is two positions pretending to be one. Split it.
Rule two, avoid overlap. "The challenge" and "the obstacle" are the same position with different words. Test every pair by asking whether a single card could plausibly answer both. If yes, merge them.
Rule three, make positions answerable by a card. "What is his phone number" is not answerable. Neither, in practice, is "the exact date this happens." Positions work when they ask about qualities, dynamics, attitudes, and directions.
Rule four, put the actionable position last. Readings that end on outcome cards feel passive. Readings that end on advice or next step cards send the reader off with something to do.
Here is a well formed five position set for a creative project:
1. What this project is really about, underneath the stated goal
2. The strength you are underusing
3. What is genuinely slowing progress
4. What the work needs from you next
5. What finishing it would change
Notice that no two positions could be answered by the same card, and the last one gives the reading a purpose.
Step Four: Give It a Shape
Layout geometry is not decoration. Physical placement affects how you read relationships between cards.
A row implies sequence. Use it for time, process, or steps.
A column implies layers, from surface at the top to depth at the bottom, or the reverse. Good for internal states.
Two columns imply comparison. Ideal for choices, or for two people in a relationship reading.
A cross or centre with satellites implies one core matter surrounded by influences. Good for situations where one thing dominates.
A triangle or pyramid implies foundation and result, with the base supporting the apex.
Choose the shape that matches the logic of your positions. If your positions are comparative and your layout is a row, readers will unconsciously look for a sequence that is not there. This mismatch is subtle and it genuinely damages readings.
Sketch the shape, number every position, and write the question next to each number. Keep this sheet, because you will forget your own positions within three weeks otherwise.
Step Five: Test, Trim, and Keep Notes
A new spread is a draft. Run it at least five times before deciding it works.
Test on known situations first. Use it on something already resolved, where you know the answer. If the spread produces a reading that matches what actually happened, the positions are doing real work. If every card feels generic, the positions are too vague.
Watch for the dead position. In nearly every first draft, one position consistently produces nothing useful. You will recognise it because you keep skipping over it or reading it as an afterthought. Cut it.
Watch for the crowded position. The opposite problem, where one position keeps needing extra explanation. That position is doing two jobs and should be split into two.
Log the readings. Date, question, cards, and a one line verdict on whether the spread helped. After five entries the pattern is obvious, and you will know whether to keep, revise, or retire the layout.
Revise on paper rather than in your head. Version two of a spread should look different from version one, with positions rewritten in clearer language. Most good custom spreads reach their final form on the third revision.
Common Mistakes and How to Fix Them
Too many positions. The classic. Cut until it hurts slightly, then stop.
Positions that ask for certainty. Anything phrased as "will this definitely happen" sets tarot up to do something it cannot do. Rephrase toward direction and influence.
Themed positions with no meaning. Assigning positions to elements, planets, or chakras looks elegant and often produces slots nobody can interpret. Only use a themed structure if each themed position translates into a real question.
No advice position. Readings that only describe leave people stuck. Add one forward looking card.
Copying a spread you have not tested. Adapting an existing layout is fine, and it still needs the five test runs.
Once your spread stabilises, name it and write it in the front of your journal with the sketch. Custom spreads become genuinely valuable in their second year, when you have twenty logged readings using the identical structure and can compare them directly. That comparability is the real payoff.
If you want models to study before designing your own, read through a structured relationship layout in our love spreads guide and a practical one in the career spread, paying attention not to the cards but to how each position is phrased.