Reading and spelling rest on two mirror-image skills. Blending is pushing sounds together into a word — /d/ /o/ /g/ becomes "dog". Segmenting is pulling a word apart into its sounds — "dog" becomes /d/ /o/ /g/. Blending powers reading; segmenting powers spelling. Together they are the bridge a child crosses from knowing sounds to actually using them.
This guide shows how to teach both, why you should start with no letters at all, and the small adjustments that fix the most common place children get stuck.
Free phonics worksheets — free
No-prep, decodable Science-of-Reading practice — print tonight, teach tomorrow.
Blending for reading, segmenting for spelling
These two skills are the same operation run in opposite directions. When a child reads, they take the separate sounds on the page and blend them into a whole word. When a child spells, they take a whole word and segment it into the sounds they need to write. A reader needs both, and they reinforce each other.
Because they are central to decoding and encoding, blending and segmenting are arguably the most important rungs of the whole sound-awareness ladder. A child can know every letter sound and still not read a word until they can blend — and conversely, blending is what makes that letter knowledge pay off.
Start with the ear, not the page
The smartest move is to teach blending and segmenting orally first — with no letters in sight. Say "/c/ /a/ /t/" and have the child tell you the word. Say "map" and have them tap out /m/ /a/ /p/. Doing this in the dark, eyes closed, isolates the listening skill from the separate challenge of recognising letters.
This matters because the most common reason a child "cannot sound out" printed words is not the letters — it is that the underlying oral blending is shaky. A week of sound-only blending games very often unlocks decoding that looked stuck. Build the ear, then bring in print.
How to teach oral blending
Keep blending playful and quick. The key technique for a child who sounds out but then cannot pull the word together is to shorten the gaps between the sounds:
- Robot talk: you say the segmented word ("get your /h/ /a/ /t/"), the child blends it to act.
- Start with two parts: blend onset and rime first ("/c/ ...at" → cat) before three separate sounds.
- Stretch, do not chop: say "mmmaaap" run together rather than "m — a — p" with big pauses.
- Continuous blending: hold the stretchy sounds and glide between them so the word almost says itself.
- Always say the word back: after they blend, repeat the whole word so the part-to-whole link is clear.
How to teach segmenting
Segmenting is the spelling-side skill, and it tends to be a little harder than blending because the child has to generate the breakdown themselves. Give them something physical to do with each sound so it is concrete:
- Tap it out: tap a finger (or a counter) for each sound in a short word.
- Push counters: slide one token into a box for each sound you hear in "sun".
- Robot voice back: you say "fish", the child answers in robot segments /f/ /i/ /sh/.
- Stretch and count: stretch the word slowly together, then count the sounds.
- Last-sound, first-sound: warm up with "what is the first sound in dog? the last?" before full segmenting.
Bringing it to print
Once a child can blend and segment by ear, the same skills transfer straight to letters. For reading, they touch each letter, say its sound, and blend across the word ("/m/ /a/ /p/ — map"). For spelling, they segment the spoken word, then write a letter for each sound. Now phonics is doing exactly what it is for.
Keep the print practice short, frequent and successful — five to ten minutes, ending on a word read or spelled smoothly. Simple consonant-vowel-consonant words (cat, sun, map) are the natural first targets, because three clean sounds are the easiest to blend and segment. A structured set of decodable practice pages gives you a ready supply of the right words so every session reinforces the bridge instead of overwhelming it.
Skip the setup work
Everything in this guide works with paper and a pencil. If you want the structure ready-made — designed, tested and printable in minutes — the matching templates are below.
Frequently asked questions
What is the difference between blending and segmenting?
Blending pushes separate sounds together into a word (/c/ /a/ /t/ → "cat") and is the skill behind reading. Segmenting breaks a word into its sounds ("cat" → /c/ /a/ /t/) and is the skill behind spelling. They are the same process in opposite directions, and a reader needs both.
My child sounds out every letter but cannot say the word. What helps?
This is the classic blending stumble. Shorten the gaps between the sounds and use continuous blending — stretch and glide ("mmmaaap") instead of chopping ("m — a — p"). Practising orally, without the letters, for a short stretch usually unlocks it, then return to print.
Should I teach blending and segmenting with letters or without?
Start without letters. Teaching these skills orally first separates the listening from letter recognition, and it is often the missing piece when decoding looks stuck. Once a child can blend and segment by ear with short words, bring in the letters and the skill transfers directly.
Sources & further reading
- National Reading Panel (2000). Teaching Children to Read — NICHD report on phonemic awareness, phonics, fluency, vocabulary and comprehension
- Gough, P. & Tunmer, W. (1986). Decoding, Reading, and Reading Disability — the Simple View of Reading. Remedial and Special Education, 7(1)
- Castles, A., Rastle, K. & Nation, K. (2018). Ending the Reading Wars: Reading Acquisition From Novice to Expert. Psychological Science in the Public Interest, 19(1)
- Ehri, L. (2005). Learning to Read Words: Theory, Findings, and Issues. Scientific Studies of Reading, 9(2)
- Foundational Skills to Support Reading for Understanding in Kindergarten Through 3rd Grade — IES / What Works Clearinghouse Practice Guide (2016)
About this guide. Published by OmniAura Digital, the team behind the printables and templates in our library. Our reading guides summarize published, peer-reviewed research (cited above) in plain English — we link the primary sources so you can verify every claim. Spotted an error? Tell us at support@omniauradigital.com and we'll fix it.