What the data shows
Between August 2025 and April 2026 there were 308,770 apprenticeship starts in England, 8.7% more than in the same nine months a year earlier. On the face of it that is a straightforward recovery story. Underneath it, the three age groups are moving in different directions.
Starts by people aged 25 and over rose 17.5%. Starts by people aged under 19 fell 5.4%. Because the two groups are moving apart rather than together, the composition of the programme has shifted: under-19s now account for 20.6% of starts, and people aged 25 and over for 52.4% – more than half.
That majority is not brand new. The 25-and-over share first passed 50% in 2020/21 (50.3%) and stood at 51.3% in 2024/25. What the current period adds is the highest share in the published series, and a continuation rather than a turning point.
This is not a one-year wobble. In 2017/18 there were 106,570 under-19 starts. In 2024/25, the most recent complete year, there were 74,990 – a fall of 29.6% across the period. Starts by people aged 25 and over over the same years went the other way, from 155,480 to 181,270.
Interpretation
Three cautions before anyone reaches for a conclusion.
First, a start is not a person. Starts count programmes, and someone who begins two apprenticeships is counted twice. The measure that counts learners is participation, which stood at 735,500 in the current period.
Second, the falling under-19 share is arithmetic, not evidence of intent. A share can fall because the numerator shrinks, because the denominator grows, or both – and here both are happening at once. The under-19 share would have fallen this year even if under-19 starts had held completely flat.
Third, the published data cannot tell you why. Employer demand, funding rules, the mix of standards approved for delivery, and what school leavers are choosing instead are all plausible contributors, and none of them is measured in this release. What the release does establish is the direction and the scale.
The same story in the participation data
Starts capture new programmes beginning. Participation counts the learners actually on a programme at any point in the year, and it moves more slowly, because it reflects several years of starts working through. It tells the same story with less noise.
Under-19 participation has fallen from 176,550 in 2017/18 to 130,840 in 2024/25. Participation by people aged 25 and over rose over the same years, from 364,650 to 377,680. Two related measures from the apprenticeship administrative data point in the same direction, which makes the compositional shift clearer than either would alone.
The 19–24 group is the quiet one in all of this. Starts stand at 83,570 in the current period, and this group has neither grown like the over-25s nor fallen like the under-19s. For anyone tracking whether apprenticeships still serve young adults leaving education, it is the group worth watching next.
Why it matters
Apprenticeships are still routinely described as a school-leaver route, and that description is doing less work every year. Around one start in five now goes to someone under 19. For a careers adviser, that is not a reason to stop talking about apprenticeships – there were still 63,530 under-19 programme starts during the current period. It is a reason to be precise about which apprenticeships are actually open to a 16- or 18-year-old, because the growth is concentrated at levels and in occupations that mostly recruit adults.
For policymakers the question is sharper. If the apprenticeship system is increasingly serving people already in work, that may be exactly what a skills programme should do – or it may mean the route out of school is narrowing while attention is elsewhere. The data cannot settle that. It can stop the argument being had with the wrong numbers.