What the law actually requires
The national curriculum for English is statutory in maintained schools in England, and only its programmes of study carry that force — the notes and guidance printed alongside them are explicitly non-statutory.
What the statutory part says is specific and quite narrow. In Year 1, pupils must be taught to apply phonic knowledge and skills as the route to decode words, and to read accurately by blending sounds in unfamiliar words containing the letter-sound correspondences they have been taught. In Year 2, they must continue to apply that knowledge until decoding is automatic and reading is fluent.
The phrase rigorous and systematic phonics programme appears only in the non-statutory notes. The word synthetic does not appear in the document at all.
The early years framework, which is statutory and in force from September 2026, works the same way: it sets out what children should be able to do rather than naming a method. Reading is described as two dimensions — language comprehension and word reading — and the early learning goal asks a child to say a sound for each letter and at least ten digraphs, to read words by sound-blending, and to read simple sentences consistent with their phonic knowledge.
So systematic synthetic phonics is a government expectation rather than a statutory form of words, and it is enforced through four separate levers: the Reading Framework, the list of validated programmes, the screening check, and inspection. A government response published in November 2025 restated the intention to reinforce that expectation, which is a fair indication that it is not about to loosen.
The Reading Framework and the validated list
The Reading Framework is guidance rather than statutory guidance. It was first published in July 2021 and substantially expanded in July 2023, when it was extended up to Year 9 and gained sections on fluency and assessment. It is the document that states in terms that teachers are required to teach a programme of systematic phonics from Year 1, and that schools should be confident in the rationale for it.
Alongside it sits a list of phonics programmes validated by the Department for Education — 45 of them, across three rounds of assessment, the list having last been updated in February 2026. Validation is worth understanding precisely, because it is routinely overstated: it is a self-assessment checked against published criteria by a panel, confirming that a programme meets those criteria. It is not a finding that one programme works better than another, and no comparative trial underlies it. Letters and Sounds, which many schools used for years, was removed from the list in 2021.
The department has said the validation process is complete and that no further round is imminent, without saying the list is permanently closed.
The screening check
The phonics screening check rests on a 2004 order made under the Education Act 2002. A child reads aloud, one to one with an administrator, twenty real words and twenty invented ones. The invented words are the point of it: a child cannot have met them before, so they cannot be recognised from memory and must be decoded.
Eligibility is defined by age and curriculum stage rather than by year group — every pupil who will have turned six and completed the first year of the key stage 1 English programme by the end of the year — though in practice that means Year 1. The check is administered in a single week in June. In 2027 that week runs from 14 to 18 June, with a variation week following.
The retake is compulsory, not optional. A pupil who did not meet the expected standard, or who has not taken the check before, must take it in Year 2.
The threshold was 32 marks out of 40 every year from 2012, when the check was introduced, until 2025. In 2026 it moved to 31 — the first change in fourteen years. The explanation given is that papers are built to the same specification each year but the words differ, so difficulty varies slightly and the mark is adjusted to hold the standard steady. It is therefore not a figure to rely on from memory; it is published in late June each year.
On the most recent published results, for 2025/26: 82 per cent of Year 1 pupils met the expected standard, against 80 per cent the year before and 58 per cent in the check's first year. By the end of Year 2 the figure was 89 per cent. Figures from 2024 onwards are not directly comparable with earlier ones, because the method of counting changed.
The check is an England-only arrangement, and that is about to stop being true. Northern Ireland will run its first Year 2 phonics check in June 2027, described by its education department as a low-stakes assessment administered by the class teacher, with results not published at school level and not used for accountability.
The two kinds of book, and the trouble with colour bands
Decodable books exist because most books are not decodable for a child who is still learning. The Reading Framework's position is that practice books should contain no letter-sound correspondences a child has not yet been taught, and should run alongside or slightly behind the phonics teaching, so that the experience is one of success rather than of guessing.
Books read aloud by an adult do a different job entirely, and are not required to be decodable. They are chosen to hold attention, to provoke a response and to extend vocabulary — all of which can happen well above a child's own reading level. Conflating the two produces the familiar scene of a child labouring through a book chosen for its story, and getting nothing from either activity.
Colour bands are more contentious than their ubiquity suggests. They originate in a handbook on book banding published by the Institute of Education at the University of London, now in its fourth edition from 2012, and came out of the Reading Recovery network — academic in origin rather than commercial, with publishers subsequently applying the scheme to their own ranges. There is no departmental, inspectorate or agency definition of a band, and no statutory national standard behind them.
The Reading Framework is in fact critical of how banding is commonly used, quoting an inspection finding from 2017 that organising books into bands hindered accuracy, and that schools mixing several schemes bought at different times failed to match books to children properly. Its own test is different and sharper: books should be ordered to follow the sequence in which the school's own programme introduces letter sounds. A book containing the word play should not come home before ay has been taught.
The other three nations
The four education systems diverge here more than almost anywhere, and treating British practice as one thing goes wrong quickly.
- Scotland. The literacy outcomes and benchmarks under Curriculum for Excellence are guidance supporting professional judgement, not a prescribed method. The word phonics appears once in the relevant outcomes. Education Scotland's own professional support material does say that explicit and systematic teaching of phonics should be a key part of how reading is taught, and that decoding should be the first approach to word recognition for most children — but it mandates no programme and does not use the word synthetic. Standardised assessments are taken in P1, P4, P7 and S3; there is no phonics check.
- Wales. Under the Curriculum for Wales, the four statements of what matters for languages, literacy and communication are mandatory, and none of them mentions phonics or decoding. The descriptions of learning, which guide progression, do: grapheme-phoneme correspondence, segmenting and blending all appear there. Personalised assessments in reading and numeracy are statutory from Year 2 to Year 9. There is no phonics check.
- Northern Ireland. The statutory minimum content for language and literacy names talking and listening, reading and writing, and does not mention phonics. In December 2025 the education department launched a literacy framework for primary schools with accompanying guidelines which are explicitly not prescriptive, but which recommend that all schools adopt a systematic phonics programme in the foundation stage. The guidelines note the evidence for synthetic phonics without naming a programme, and add that phonics is one part of a rich literacy environment rather than the whole of it.
What the evidence supports, and where the argument actually lies
The Education Endowment Foundation's assessment of phonics is unusually strong by the standards of educational research: an average of five additional months' progress, at very low cost, on evidence rated high and drawn from 228 studies. Its summary wording is "high impact for very low cost". Anyone describing the evidence base as thin is not describing this literature.
The important caveat comes from the same source and is frequently left off. Phonics improves the accuracy of a child's reading, and not necessarily their comprehension. Where a child's difficulty lies in vocabulary or in understanding what has been read, other approaches are likely to be more use, and phonics is explicitly described as one component of a literacy strategy rather than the strategy itself.
That is also where the genuine dispute sits. Researchers at the UCL Institute of Education argued in a paper publicised in January 2022 that England had moved from a balanced approach to a narrow one, that the policy was not underpinned by the most robust available evidence, and that trials they reviewed from Canada and Norway found the most effective teaching combined phonics with work on whole texts. Their survey of 2,205 teachers found that of 936 written comments about the screening check, all but one were negative, with the word pressure appearing 97 times and many teachers describing additional lessons laid on to get pupils through it. More than 250 signatories wrote to the then education secretary asking for greater reliance on teacher judgement. The authors acknowledged they might have missed relevant longitudinal trials.
The government's position, restated in November 2025, is to reinforce the expectation of a systematic synthetic phonics programme in all primary schools, with an ambition that 90 per cent of children meet the expected standard in the check. Set against that, an inspection report from October 2024 found reading teaching improving while English curriculums were often weak beyond phonics — which is close to the criticism, arriving from the other direction.
What a parent can do that is known to work
The evidence on home involvement is good, and it does not require buying anything.
- Reading aloud, with talk attached. Approaches built on communication and language in the early years show around seven additional months' progress on moderate evidence from 66 studies, and the consistent finding is that reading to a young child works when the child is encouraged to answer questions and discuss the story, rather than listen in silence.
- Taking the two books seriously as two jobs. The practice book is for accuracy and fluency, read more than once, and the same text reread is doing exactly what it is supposed to do. The read-aloud book is for everything else.
- Parental engagement generally shows about five additional months' progress on extensive evidence from 94 studies — with the honest qualification that many of those studies were not independently evaluated.
- The library. In England and Wales, free lending of printed material is a statutory requirement rather than a local courtesy, under the Public Libraries and Museums Act 1964. Libraries are devolved elsewhere: in Northern Ireland free membership is policy, with under-16s joining at a branch with a parent or guardian. One caution worth stating, because it is widely assumed otherwise — there is no statutory exemption for under-18s from overdue charges. Whether a child is charged is a decision of the individual library service, and practice varies, so it is a question to ask locally rather than assume.
- The Summer Reading Challenge, run by The Reading Agency with public libraries, is free and aimed at ages four to eleven, and runs each summer.
- Bookstart packs, from BookTrust, are gifted to young children in England, Wales and Northern Ireland, with free library sessions for under-fives attached.
2026 was designated a National Year of Reading, launched in January of that year by the education department with the National Literacy Trust, built around a request that families read together for ten minutes a day. The figures used to justify it are worth recording: only one child in three aged 8 to 18 said they enjoyed reading in their spare time, the lowest since 2005, and more than a quarter leave primary school below the reading age expected of an eleven-year-old.
What changes next
A review of the curriculum and assessment reported in November 2025, and the revised national curriculum arising from it is due to be published by spring 2027, for first teaching from September 2028. Two assessment changes are intended alongside it: a statutory Year 8 reading test covering fluency and comprehension, and strengthened writing assessment at the end of Year 6. No start date has been announced for either.
Where reading difficulty looks like it needs more than home support, the evidence on what actually helps is set out under hiring a tutor, and the short answer there is that the school should be asked first.