Foundations
What Order to Teach Letters
There is no single research-proven order for teaching letters. Here is the sequence we use, why we chose it, and a five-minute way to assess any program.
No letter order has been proven best by research, so the honest question is not which order is correct but what an order is designed to let a child do first. Ours is built for a parent teaching at home in a few minutes a day, so it is arranged to reach a readable word as fast as possible and to keep look-alike letters apart.
We teach letters in this order:
The order Wonster Words teaches the twenty-six letters in: a, t, b, l, i, f, p, s, o, m, r, u, h, c, g, d, n, j, v, w, e, x, k, y, z, q.
VowelConsonant
A child who knows the first 4 can already read at, bat and tab. A child working alphabetically is still on d at that point, with nothing to read. That single difference is most of the argument, and the rest of this page is the reasoning, the numbers behind it, and a five-minute check you can run on any program’s order, including ours.
Letter order is one rung of a longer ladder. The stages either side of it are in what order to teach phonics skills, and this page assumes you already know how to teach one letter: the name, the sound, and how to say it without an “uh” on the end.
Does the order you teach letters in actually matter?
The order matters less than programs selling you one would like you to think. Literacy researcher Timothy Shanahan put it plainly in Putting on Your Underwear First: “there is no research-proven best sequence for teaching the ABCs or phonics.”
The evidence for that is a collection of programs that all worked despite disagreeing with each other. Writing about the studies the National Reading Panel gathered, Shanahan notes that “about 18 different phonics curricula were examined in that collection of studies, and each had its own sequence for introducing letters and sounds. And they all worked.”
That is a concession, and it cuts against us as much as anyone. The National Reading Panel reviewed 38 studies yielding 66 comparisons and found that teaching phonics in a planned sequence works significantly better than teaching it as words happen to come up, while also finding no single best approach, because children differ in what they arrive with. The finding is that having an order is beneficial, but the exact order is less important.
What the Panel did document is that programs differ in how they sequence instruction, with some teaching most letter-sounds before children read any words and others starting word reading much sooner, and both kinds were represented among the programs it found effective.
“No proven best” is not the same as “it makes no difference.” It means no sequence wins for every child in every setting, which leaves a narrower and much more useful question: best for whom, doing what, with how much help? A sequence built for a kindergarten class of 30, taught five mornings a week by someone trained in it, is solving a different problem from a sequence a parent works through on a sofa in five minutes before bed. Both can be right. Neither is right everywhere.
Ours is built for the second case, and the rest of this page is what that choice commits us to.
If no order is proven best, what should an order be optimized for?
If no sequence wins in the abstract, what a sequence is built to achieve is all that is left to ask, and almost every published order was built for a classroom. SATPIN, the starting set used by Jolly Phonics and much of the synthetic phonics world, assumes a trained teacher, a group of children, five sessions a week and a scope and sequence to fall back on. Under those conditions you can spend the first two weeks on letter-sounds before any word gets read, because attendance is not in question.
Home is a different set of constraints. There is one adult, usually untrained, working in the gaps of a day, and a child whose attendance is entirely voluntary. A session that ends with nothing read is a session a 4-year-old has no reason to repeat. So the thing to optimize for at home is the distance to the first win, and after that, everything that stops a parent having to notice and correct a mistake in the moment.
The National Reading Panel grouped systematic phonics into three families and found benefits across all of them: synthetic programs, which convert letters to sounds and blend them; larger-unit programs, which work with chunks of words such as the rimes in word families; and a miscellaneous group. Ours is a larger-unit program. Most of the ones you will compare it against are synthetic. Neither is the correct answer.
What order does Wonster Words teach letters in, and why?
Four rules built this sequence, and between them they decide every position in it.
The order Wonster Words teaches the twenty-six letters in: a, t, b, l, i, f, p, s, o, m, r, u, h, c, g, d, n, j, v, w, e, x, k, y, z, q.
VowelConsonant
1. Start with /a/, the vowel that builds the most words
A sequence with no vowel builds nothing, so ours opens with one, and it opens with the most productive one. Short a is the busiest vowel in short words: among the three-letter consonant-vowel-consonant words in Webster’s Second, 217 use a, ahead of o at 169, e at 137, i at 136 and u at 113. It also yields the most word families with three or more members to swap between: 17, against 16 for e and 11 for u.
It is also the easiest to say. Short a as in cat is a single, stable, open-mouthed sound that does not shift much between accents, which matters when the person modeling it is a parent rather than a trained teacher.
2. The first consonants are the ones whose name tells you their sound
A child who knows a letter’s name is halfway to its sound for most letters, and nowhere near it for six of them. Say bee, tee, pee, ef, el, ess and you have said /b/, /t/, /p/, /f/, /l/ and /s/ on the way. Say aitch, double-u or wye and you have said nothing like the sounds h, w and y actually make. The same goes for cee and gee, whose names carry the soft sounds while early reading needs the hard ones.
All six of those letters sit at position 13 or later: h, c, g, w, y, q. Every consonant in the first 12 has its sound audible in its name. That is a checkable claim about the sequence above, and it is why a parent teaching the early letters rarely has to explain that the name is a red herring.
3. Build toward real words as fast as possible
Each early letter is chosen for what it makes with the letters already taught, not for how common it is in general. a and t give at. Adding b gives bat and tab. l gives lab and completes the first family worth playing with. i, the second vowel, opens it, bit and lit at the 5th letter.
That is the metric we designed against, and the sections below are what it buys against every other sequence we could compare.
4. Spread the vowels out, and keep confusable letters apart
Vowels arrive at 1, 5, 9, 12 and 21 rather than in a block. Four of the five are in by the 12th letter, so a child is reading across different vowel sounds early instead of doing nothing but -a- words for a month.
The exception is e, held back to 21 on purpose, 16 positions from i, because short e and short i are the pair children confuse most. The same thinking keeps b and d 13 apart, p and q 19 apart, and n and u 5 apart. n and u matter to us specifically: they are a rotation of each other, and the app ships a lesson for that exact confusion.
How soon can a child read a real word?
A child can read a word by the 4th letter, and by the 5th they have two word families to play with. With a, t, b and l they can read at, bat, tab and lab. Add i and it, bit and lit arrive. That matters more than it sounds, because it changes what a practice session is: instead of ten minutes of letters that do not do anything yet, it is a minute of letters and then reading a word.
Counted properly, the first 4 letters unlock 22 regular words a beginner could sound out, which is more than any sequence we compared. Hooked on Phonics and Reading.com reach 19, Reading Eggs and Teach Your Monster 18, Jolly Phonics 16, and alphabetical order 12. The method behind those counts is further down, and it is worth saying plainly that most of the 22 are dictionary words rather than words a 4-year-old wants: the everyday handful is at, bat, tab, lab, all, ball and tall.
Why do word families come before sounding out?
Rhyme develops before a child can pull a word apart into separate sounds, so a sequence that reaches a word family early is working with that grain rather than against it. A 3-year-old who can tell you that cat and hat go together usually cannot yet tell you that cat is /c/ /a/ /t/. Word families sit on the first ability. Sounding out sits on the second.
This matters at home more than it does in a classroom, because blending is exactly where home practice stalls. It is the single most common thing parents describe: the child knows every letter sound, says all three correctly, and then says a different word or nothing at all. If the first readable word depends on blending, the first win waits behind the hardest step, and for some children it waits months.
Starting with a family moves the win in front of that step. A child reading bat and then tab by swapping the front letter is reading, and they are doing it with a skill they already had. The families themselves, -at and -it among them, are on our CVC word list. Sounding out arrives next and it still has to arrive: word families are efficient and they are also a trap, because a child can get very good at swapping the first letter without ever learning to sound a word out. That is why what to do when a child cannot blend is the next thing to read, and why teaching rhyming is worth doing before any of this.
Should vowels come first, and why is e last?
Every word needs a vowel, so a sequence that saves them up has nothing to read until it stops. Ours opens with one. The second arrives at 5, the third at 9, the fourth at 12 and the last at 21.
The last one is e, and the distance is the point. Short e and short i are the two vowels children confuse most, because the sounds sit close together and the mistake produces a real word rather than a nonsense one: a child reading pen as pin has read something, so nothing about it sounds wrong. That is 16 positions between i and e, the widest gap in any sequence we compared. Hooked on Phonics has 14, IMSE 9, Reading.com and Teach Your Monster 8, Reading Eggs 7, alphabetical order 4 and Jolly Phonics 3.
The cost is real and worth naming: holding e back until position 21 means bed, red and ten are late words in this order, and Jolly Phonics, Reading Eggs, Teach Your Monster and Reading.com all make the opposite trade.
Which letters do children confuse the most?
Reversals are not a sign of anything wrong, and the sequence can prevent most of them for free. b and d are the same shape facing opposite ways, and a child learning both in the same two weeks has to hold that difference before they have much reason to care about it. This is the one part of sequencing Shanahan is prescriptive about: “Don’t teach b and d together, or m and n, for instance. Letters that are visually or phonemically similar need to be kept apart.”
In this order b sits at 3 and d at 16, 13 positions apart, and p and q are 19 apart. Both are the widest of any sequence we compared:
| Sequence | b to d | p to q |
|---|---|---|
| Wonster Words | 13 | 19 |
| Reading Eggs | 12 | not taught apart |
| Teach Your Monster | 9 | not taught apart |
| Reading.com | 9 | not taught apart |
| Hooked on Phonics | 8 | 3 |
| IMSE (Orton-Gillingham) | 7 | 13 |
| Jolly Phonics | 6 | not taught apart |
| Alphabetical | 2 | 1 |
The reason this is a home argument rather than a general one: in a classroom a teacher sees the reversal happen and corrects it in the moment. At home, nobody is watching every attempt, and a mistake made ten times unnoticed is harder to undo than one caught the first time. The sequence does the work the supervision would have done.
Three pairs are worth naming as costs rather than wins, because a sequence cannot separate everything. Shanahan names m and n, and this order keeps them only 7 apart, which is mid-pack among the eight sequences in the table above. o and u sit 3 apart, which is close for two short vowels that already sound alike in hot and hut; that is the direct price of getting four vowels in by the 12th letter, and we would make the trade again. And l and i are adjacent at 4 and 5, which matters not for their sounds but because a lowercase l and a capital I are near-identical in many typefaces.
Whether that last one bites depends on the letterforms a program uses rather than on the sequence, so it is worth looking at how your program draws them. It is the one place where our own order is worse than the alphabet.
How long should you spend on each letter?
A week per letter is a long time to spend on something a child can learn in a few minutes, and it is the pace most preschool programs run at. Cindy Jones and Ray Reutzel tested the alternative across 13 kindergarten classrooms and reported it in Reading Psychology in 2012: brief lessons, more exposure to each letter, extra attention on the letters children find hardest, and letters returning in repeated cycles of review rather than being met once and left behind. Children learned their letters more successfully that way. It was a classroom study rather than a home one, but the part that transfers is the pace.
Twenty-six days at a letter a day still leaves room to revisit every one of them inside a term, which letter-of-the-week does not. What a session itself looks like is a question about teaching a letter rather than about ordering them: how many minutes, how many letters at a time, what to review before adding a new one.
Software solves half of this. An app can present the letters in a chosen order and keep the practice short, and that is most of the benefit. Bringing a specific letter back at the right moment is something you are still doing yourself, by paying attention to which ones are shaky. No app we know of, ours included, removes that part.
Is SATPIN the best set of letters to start with?
SATPIN is the best-known starting set in phonics: s, a, t, p, i and n, the six letters most English school programs open with. They are chosen on much the same logic as ours, which is that they combine into short words quickly, and it works. The difference is what comes next and what the sequence is willing to trade.
It reached English classrooms through Letters and Sounds, the Department for Education’s 2007 phonics framework, which opened with satpin and became close to a national default. Two things about that are worth knowing before you treat it as settled. The DfE removed Letters and Sounds from its list of validated phonics programs in 2021 (with schools given until 2022 to move), though for reasons about the handbook not being a complete program with training and matched books, rather than anything about the order itself.
And satpin is neither British nor modern in origin. Gill Cochrane and Greg Brooks traced it in the British Educational Research Journal in 2022 back to 1960s American Orton-Gillingham practitioners, Sally Childs and Aylett Cox, whose 1967 rationale for a version of satpin they conclude “needs updating”. A sixty-year-old starting set that spread because a national framework used it is not the same thing as a proven one, which is the point this page opened with.
The clearest way to compare two sequences is to ask what a child can read at the same point in each. After 4 letters, a child using our sequence can read:
| Sequence | Words readable after 4 letters |
|---|---|
| Wonster Words | 22 |
| Hooked on Phonics | 19 |
| Reading.com | 19 |
| Reading Eggs | 18 |
| Teach Your Monster | 18 |
| Jolly Phonics | 16 |
| IMSE (Orton-Gillingham) | 13 |
| Alphabetical | 12 |
The 4-letter checkpoint decides whether a child’s first week produces anything, which is why it is the one drawn. It is also the one we win by the widest margin, so here is every checkpoint, including the ones we lose:
| Sequence | 4 letters | 6 | 8 | 10 | 12 | 15 |
|---|---|---|---|---|---|---|
| Wonster Words | 22 | 56 | 138 | 280 | 516 | 993 |
| Jolly Phonics | 16 | 63 | 155 | 314 | 526 | 1,128 |
| Reading.com | 19 | 65 | 151 | 280 | 455 | 918 |
| Reading Eggs | 18 | 64 | 140 | 282 | 414 | 953 |
| Teach Your Monster | 18 | 63 | 140 | 282 | 414 | 953 |
| Hooked on Phonics | 19 | 52 | 128 | 260 | 448 | 898 |
| IMSE (Orton-Gillingham) | 13 | 40 | 95 | 213 | 279 | 609 |
| Alphabetical | 12 | 21 | 52 | 89 | 206 | 544 |
Read the ends of that table rather than the middle. We lead early and Jolly Phonics leads from 12 letters on, which is where a synthetic program is designed to win. That is the trade, stated plainly: we buy the first word and give up some of the later volume. If your child is in a school using satpin, that is not a reason to switch either way.
How the counts were made, so the table can be checked rather than believed:
the list is 6,429 two-to-four-letter words from Webster’s Second, filtered to the
regular ones a beginner could decode (a single vowel letter, one of the patterns VC, CVC, CCVC,
CVCC or VCC, no silent e and no q), which leaves 3,014 words. A word counts at the
point where every letter in it has been taught. The word list is the one shipped as
/usr/share/dict/words on macOS, so the whole table can be reproduced in a few
lines of code.
How can you tell if a program’s letter order is any good?
Every check below can be run on a published scope and sequence, including ours, and none of them needs a teaching qualification. If a program will not publish its letter order at all, that is the first answer.
Find a vowel in the first two letters
Read the first two letters of the sequence. At least one should be a vowel. Without a vowel there is no word to make, so a child works through letters with nothing to show for them.
Check that the early letters are common ones
The first six or eight letters should be letters that turn up constantly in short words, such as a, t, s, m and p. A sequence that opens with j, v or z is optimized for something other than reading words.
Measure the gap between b and d, and between p and q
Count the positions between b and d in the sequence, and between p and q. The wider the gap, the less likely your child is to be learning two mirror-image letters in the same two weeks.
Check how long the capitals wait
Every good sequence starts with lowercase. Look at how soon the capital for each letter arrives after it: paired with its own letter, or held back as a separate block after all twenty-six. The print a child meets outside a lesson, on signs, packaging and their own name, is full of capitals.
Ask when the first real word appears
Work out how many letters a child learns before the sequence lets them read a whole word. If the answer is more than four or five, ask what the child is doing in the meantime.
Run them on this page’s order and it passes four cleanly and makes a deliberate trade on the fifth: e arrives late, so a few common words arrive late with it. Check 4 is the one whose reasoning is not on this page, because it is about the letters rather than their order: how long the capitals should wait. A program that passes all five and tells you its sequence is proven is still overselling, because no sequence is.
If you are choosing between apps rather than sequences, how to choose a reading app and the phonics app comparison go wider than letter order alone.
Frequently asked questions
What is the best order to teach letters for reading?
There is no order that research has proven best, and programs that claim one are overselling. What good orders have in common is that they put a vowel in early, start with letters that appear in many short words, keep letters that look alike far apart, and let a child read a real word within the first few lessons rather than after all twenty-six. Our order is a, t, b, l, i, f, p, s, o, m, r, u, h, c, g, d, n, j, v, w, e, x, k, y, z, q.
Why teach letters in a non-alphabetical order?
Because alphabetical order delays the first word. A, b, c, d and e give a child almost nothing to read, so several lessons pass with no payoff. An order chosen for reading front-loads the letters that combine into short words, which means a child can read something within the first few days. The alphabet song is for naming letters in sequence, which is a different job from reading.
Is it better to teach vowels or consonants first?
Neither on its own. Mix them, and put a vowel first. A vowel with no consonants makes no words and consonants with no vowel make no words either, so the useful question is how soon the sequence gives your child one of each. Ours opens with a vowel, adds a second by the 5th letter, a third by the 9th and a fourth by the 12th.
What letters should I teach my 4-year-old first?
Start with the letters in their own name, then move to a sequence built for reading rather than for the alphabet song. If you want somewhere concrete to start, a, t and b give you "at", "bat" and "tab" almost immediately, which is the point.
Which letters are easiest for children to learn first?
The ones a child already meets count for a lot, starting with the letters in their own name. After that the best guide is the letter's own name. Say bee, tee, ef or ess and you have said the sound on the way, while aitch, double-u and wye tell a child nothing about the sounds they stand for. Those last ones are where the extra repetition goes. Sounds you can hold out, like the ones s, m and f make, are not easier to learn than the rest, but they are easier to blend, which is a different job and it arrives later.
What if my child's preschool teaches A to Z?
It is not worth a fight, and the two do not cancel out. Preschool letter-of-the-week work is mostly letter names and letter shapes, which is useful. You can run a reading order at home alongside it without confusing your child, because you are teaching the sounds rather than re-teaching the names. If anything, the overlap helps.
Can we switch to a different letter order partway through?
Yes, and you do not start again. Keep the letters your child already knows and slot the rest in using the new order, skipping what is already solid. Children do not hold an order in their heads, only the letters, so a change of sequence costs a parent some planning and costs a child nothing.
Related guides
Last updated
