School
1 in 5pupils with dyslexia reach grade 5 in GCSE English and maths
More than halfwith no recorded special educational needs
78%say people assume they aren't clever
65%hide their dyslexia
Dyslexia in Defence evidence
The data behind the story
What people across Defence are telling us, and what the research says.
What Defence is doing well
Strength in numbers: when you share your data, we can help more people.
More than 2,000 people have joined the Defence Dyslexia Network. 228 have told us how accessible Defence training really is. 92 have written, in their own words, what they want people to understand. Every answer is evidence, and evidence is where help starts. If you can't measure it, you can't improve it.
2,000+
Defence Dyslexia Network members
228
training responses
92
written lived experiences
Defence Dyslexia Network, membership at March 2025; training accessibility feedback form, 2026; lived experience responses, 2026.
People want to help, they just need the right guidance.
In a 2022 University of Southampton study of 151 Defence personnel, 58% said they would know how to help someone with dyslexia, and 108 described exactly what they would do. Their most common request was simple: training, guidance and awareness. Instructors are already making a difference. About one in five comments on what worked well in training named them.
Would know how to help: 58%
Described how they would help: 108 of 151 (72%)
About 1 in 5 training comments named instructors or staff.
Beardsell, University of Southampton, 2022 (151 respondents, voluntary survey, mostly Army). Instructor figure: Dyslexia in Defence reading of 206 free text comments on the training accessibility form (about 45 named instructors or staff).
Where needs are discussed, accessibility generally follows.
Of the 172 people who answered, 82% rated a course fully or mostly accessible when learning needs were discussed at the start. Where they weren't, 25% did.
Needs discussed at the start: 82% (105 of 128)
Needs not discussed: 25% (11 of 44)
Defence Dyslexia Network training accessibility feedback form, 2026. Self reported. Responses that did not answer yes or no are excluded.
What can be improved
From training rooms to the transition out of service, and into the wider world of school and work.
Support exists on paper. Experience depends on who you meet.
Only 6% of Defence personnel surveyed say dyslexia is fully accommodated in training and events. 68% say getting an adjustment depends on leadership discretion. 46% have felt excluded at a training event at least once. 51% aren't sure the policies that protect them, JSP 822 and ACSO 3226, even exist. And nearly one in five wouldn't know how to help a colleague.
Say dyslexia is fully accommodated: 6%
Say adjustments depend on leadership discretion: 68%
Have felt excluded at a training event at least once: 46%
Not sure JSP 822 and ACSO 3226 exist: 51%
Would not know how to help a colleague: 18%
Bars show the share of respondents. For the first bar, a low figure is the concern.
Davis, Falmouth University, 2025 (115 respondents); Beardsell, University of Southampton, 2022 (151 respondents). Self reported surveys.
Too many find out too late.
Nearly half of the neurodiverse veterans in a Cranfield University study (74 of 154) were diagnosed only after leaving service. Dyslexia was the most common condition in the study, at 32%.
- Diagnosed after leaving: 48%
- Diagnosed during service: 27%
- Diagnosed before joining: 25%
The study covered neurodiverse veterans across several conditions, not only dyslexia.
Allen et al., International Journal of Human Resource Management, 2025.
NATIONAL FIGURES, MOSTLY ENGLAND. NOT DEFENCE DATA.
Beyond Defence: it starts at school and follows people into work.
In England, only 1 in 5 pupils with dyslexia reaches a grade 5 in GCSE English and maths, against more than half of pupils with no recorded special educational needs. 78% of dyslexic teenagers say people assume they aren't clever, and 65% hide their dyslexia. Children in households earning under £30,000 are less than half as likely to be diagnosed as those earning £100,000 or more (43% against 90%). Then comes work: 81% of 16 to 24 year olds find job adverts hard to read, 87% find applications hard, and 77% worry that telling an employer will cost them the job.
School, diagnosis, work
Diagnosis
43%of children in households under £30,000 are formally diagnosed
90%in households of £100,000 or more
Work
81%find job adverts hard to read
87%find applications hard
77%worry that telling an employer will cost them the job
British Dyslexia Association, Set up for somebody else, 2025; A pathway interrupted, 2026. The school figures are for England, 2023 to 24. Other figures are UK surveys of young people.
You said. We heard. We're doing.
You said
Only 6% find training fully accommodated, 68% say it depends on leadership discretion, and 51% aren't sure the policies exist.
We heard
People don't know what they're entitled to, or who to ask.
We're building
A plain English guide to JSP 822, ACSO 3226 and your entitlements, with step by step help on asking for a diagnosis and a clear list of who to contact.
You said
Where needs are discussed at the start, accessibility generally follows.
We heard
The most powerful change costs nothing: ask.
We're doing
Briefings and support sessions for organisations, so asking becomes the first thing instructors and managers do.
You said
Too many find out too late, and 77% worry that telling an employer will cost them the job.
We heard
People need a route in, and someone who has been there.
We're building
A route to diagnosis and assessment, interview disclosure guidance, peer conversations with veterans, and CV and confidence support.
Strength in numbers.
Share your story, join the network, or partner with us.
Sources and method
Every figure on this page comes from one of the sources below. Figures are as at October 2026.
Our own data
- Defence Dyslexia Network membership: more than 2,000 members at March 2025.
- Training accessibility feedback form: 228 responses, March to September 2026. Voluntary and self reported. Most responses came from the RAF (54%) and Phase 2 training (63%), so the results may not represent all of Defence.
- Lived experience form: 92 written responses, January to February 2026.
Results are shown as totals only. No names or identifying details are published.
Research we draw on
- Beardsell, University of Southampton, 2022. Dyslexia support in the military. 151 respondents, voluntary survey, mostly Army. Used with the author's permission.
- Davis, Falmouth University, 2025. Research project on dyslexia in Defence training and events. 115 respondents. Used with the author's permission.
- Allen et al. (opens in a new tab), International Journal of Human Resource Management, 2025. Neurodiverse UK military veterans, a Cranfield University study. 232 participants, with diagnosis timing reported for 154.
- British Dyslexia Association, Set up for somebody else, 2025. A survey of 500 young people aged 13 to 18, with school attainment data for England, 2023 to 24.
- British Dyslexia Association, A pathway interrupted, 2026. A survey of 257 young people aged 16 to 24.
We summarise these reports and point to them. We do not reproduce their charts or long passages.
The 18,205 figure
18,205 is an estimate. It is 10% of 182,050 serving personnel (MOD statistics, 1 January 2026), using the British Dyslexia Association estimate that about 1 in 10 people are dyslexic. It is not a count of diagnoses.
How to read the figures
- Surveys are voluntary and self reported. People who chose to answer may differ from those who did not.
- Each source asked different people, at different times, in different ways. We do not add figures from different sources together, and some people may appear in more than one.
- "Where needs are discussed, accessibility generally follows" shows an association. It does not prove that one causes the other. It uses the 172 people who answered yes or no.
- The "about one in five" figure for instructors is our own reading of 206 free text comments. It is not a field on the form.
- National figures are for England or the UK. They are not Defence data.
- Percentages are rounded.
Spotted an error, or want to talk about the data? Get in touch.
What to do next
Pick the path that fits where you are right now.