Behind a locked door in Oxford, freezers hold blood, tissue, saliva and urine donated by people taking part in trials of treatments for mental illness. Every sample in a facility of this kind has to be traceable, from its place on the shelf back to the consent given by the person who provided it. Nothing about this looks like reform. It is the kind of detail a pharmaceutical sponsor examines when deciding where to place a study, and recruitment into commercial mental health trials in the NHS fell by 44 per cent over five years, according to one recent study.

Oxford Health NHS Foundation Trust announced this week that it has secured a research licence from the Human Tissue Authority for its Oxford BioStorage Service, which opened a year ago within the trust's NIHR clinical research facility. The Authority regulates the storage and use of human tissue for research across England, Wales and Northern Ireland, and a licence is the legal precondition for holding such material. The trust describes the licence as a framework on which to build. That is an accurate and appropriately modest description. A licence grants permission. Demand has to be earned separately, and it is the more consequential matter.

Mental illness accounted for only 3.3 per cent of the more than 4,000 interventional medicines trials approved between 2019 and 2023, a small share set against the burden of disease. The University of Manchester study behind the 44 per cent figure found sponsors increasingly choosing contract research organisations that own and run their sites. The preference is rational. Such organisations can recruit more participants from fewer sites and offer a single, predictable standard of compliance. An NHS mental health trust that wants to compete has to match that predictability, and a licensed, inspected sample store is one of the few ways to demonstrate it.

Government has spent the period since spring 2025 attacking the other end of the problem. Average set-up time for commercial trials fell from a baseline of around 250 days to 122 in the six months to March 2026, against a target of 150, and the latest quarterly figure was 100 days. The improvement is real, though that quarter rests on only 24 commercially sponsored trials, which is a thin base for declaring the problem solved. Ministers have backed the effort with £45 million for set-up capacity in NHS organisations and £92.5 million for 35 commercial research delivery centres, and the Life Sciences Sector Plan aims to quadruple commercial trial recruitment by 2029.

Set-up speed measures contracts, approvals and site agreements. It says little about whether a site can hold and trace the biological material that modern psychiatric trials increasingly depend on. The government's mental health programme, funded through the Office for Life Sciences, carries an explicit ambition to make Britain a centre for precision psychiatry, which means well-characterised cohorts with genomic and other biological data linked to stored samples. Those cohorts cannot exist without lawful storage. Clinicians in the field have also argued that mental health trusts, being organisationally separate from the acute hospitals where research is concentrated, have been poorly placed in competitions for research infrastructure funding. A licence at a mental health trust therefore fills a gap that national targets assume has already been closed.

The cost of that assurance falls on small teams. Licensing of this kind demands documented governance, trained staff and regular inspection, and those obligations land on research and development functions that are modest by the standards of the organisations that host them. Trust boards working within tight financial limits and managing demand in mental health services will be asked to keep funding a capability whose returns arrive slowly, through commercial income and national research funding that sit outside the clinical budget. Research is easy to defer when operational demand tightens and difficult to rebuild once sponsors have gone elsewhere.

For trust executives, the licence is an asset only if it is attached to a pipeline of named sponsors and a plan for throughput that survives a bad winter. Sponsors will reasonably ask for inspection outcomes and sample volumes before they commit, and trusts should expect to be transparent about both. Government should report the mental health share of commercial trial activity alongside headline set-up times, since an average across every speciality can improve while one of them stays behind. Patients in mental health services have the most to gain from trial sites close to where they receive care, and they are the group least likely to see the headline figures change anything unless sponsors are persuaded.

Oxford is an unusual place to test the proposition. It hosts a clinical research facility dedicated to mental health research and has academic relationships that most mental health trusts cannot match, so the licence may prove easier to use there than it would be elsewhere. A licensed freezer shows that one trust can hold samples to a standard sponsors recognise. A national system in which any mental health trust with a viable cohort can reach a sponsor on equal terms remains to be built. The first measure of progress will be the next sponsor that places a study in Oxford, and the one after that.