LMS Joint Research Groups in the UK (Scheme 3), 2026 Round: Up to £2,400 a Year to Run a Rotating Seminar Series Across Three or More Institutions, Closing 30 September 2026

LMS Joint Research Groups in the UK (Scheme 3), 2026 Round: Up to £2,400 a Year to Run a Rotating Seminar Series Across Three or More Institutions, Closing 30 September 2026

There is a specific kind of UK mathematics community that has no natural home in the research council system: four or five people working on the same problems at four or five different universities, none of whom has enough critical mass locally to sustain a seminar, all of whom would benefit enormously from meeting twice a year. A standard responsive-mode grant is the wrong instrument — the money needed is train fares and a room, not a postdoc. A conference grant is also the wrong instrument, because the point is regularity, not a one-off event.

Scheme 3 is the London Mathematical Society’s answer. It gives a named grant-holder up to £2,400 to run a Joint Research Group — a set of open, rotating meetings involving mathematicians at three or more separate locations who share a research interest. The 2026 application deadline is 30 September 2026, with decisions in October.

The sum is small. The function is not. A Scheme 3 grant is what turns an informal mailing list into a standing UK subject community with a name, a calendar, and a route for PhD students to meet people outside their own department. Many of the UK’s long-running specialist seminar series have been carried by this scheme for years.

Key details

What the grant actually pays for

The scheme funds the running costs of meetings, not people’s time. The LMS lists allowable costs as travel, accommodation, subsistence, caring costs, and online support — the last category explicitly including software licences and subscriptions and, where genuinely necessary, tech support hardware. That last item matters more than it looks: it means a group that wants to run hybrid meetings so that members with caring responsibilities or long journeys can join remotely can put the cost of doing that properly into the budget rather than improvising with a laptop on a chair.

The money normally covers the expenses of speakers and participants. It cannot be used to pay speakers’ honoraria. That is a firm rule, not a preference, and a budget line labelled “speaker fee” is the fastest way to have a section of your request struck out.

Two other financial constraints shape the budget. First, the £160 per day cap on accommodation and subsistence. In most of the UK that is workable; for a meeting in central London or Oxford in term time it is tight, and it is worth checking room rates before committing to a venue rotation that assumes overnight stays. Second, the grant must be claimed within the same financial year in which it is awarded, and any surplus must be returned to the Society promptly and may not be repurposed. A Joint Research Group is therefore a rolling annual commitment: you apply, you spend within the year, you report, and you apply again.

One point of caution on the headline figure. The dedicated Scheme 3 page states a maximum award of £2,400. The Society’s general grant-schemes guidance notes state £2,100 for the same scheme, and the deadline calendar page lists “up to £1,500 per academic year”. These three published figures do not agree. The scheme page is the most detailed and specific source and is the one used here, but the LMS is also clear that the published maximum is an upper limit rather than a standard award. Before you finalise a budget, email [email protected] and get the current ceiling confirmed in writing — it takes one message and removes the risk of building a meeting programme around a number that is £900 too optimistic.

Who is eligible, and who is not

Any mathematician based in the UK, the Isle of Man or the Channel Islands may apply. There is no career-stage restriction on the grant-holder beyond one exclusion: PhD students cannot apply. Students who want to organise events for other students are directed to Scheme 8, the Postgraduate Research Conference Grant, instead.

The application is made by a single nominated grant-holder who is responsible for how the money is used. That person must have the support of collaborators at at least two further institutions, so the minimum viable group is three nodes. At least two of those must be in the UK, Isle of Man or Channel Islands. Beyond that, international nodes are welcome, and the LMS makes a specific invitation on this point: applicants “might wish to include nodes from countries where mathematics and/or mathematicians are under threat”. That is an unusual and deliberate piece of guidance, and a group that acts on it should say so plainly in the academic case rather than leaving the reviewer to infer it.

If the grant-holder is not an LMS member, the application must be supported by a current member. Membership is not otherwise required, but in practice a group whose entire membership sits outside the Society will find the process easier if at least one node joins.

The Society funds mathematical activity only. Its guidance is explicit that “mathematical” here excludes mathematics education. A group whose meetings are principally about pedagogy should look at the LMS grants to support higher education teaching and scholarship instead.

The structure the LMS expects of a group

Scheme 3 is prescriptive about the shape of the activity in a way that rewards reading the criteria carefully.

Genuine multi-site involvement. All three or more institutions must be really participating. The Society “normally would expect meetings to be held at each of the institutions in rotation”. A group that in practice always meets at the grant-holder’s department with two colleagues travelling in is not what the scheme is for, and a proposed venue rotation is the cheapest way to demonstrate that you understand this.

Open meetings. Meetings are expected to be open to all. This is not a closed working group.

Active recruitment of early-career mathematicians. The Society expects meetings to be advertised to research students and young postdoctoral mathematicians, and that “every effort will be made to encourage participation by them”. Groups that publish dates well in advance, offer a slot for a student talk, and reimburse student travel from the grant are doing exactly what the scheme was designed to produce.

At least two meetings a year. The Research Grants Committee dropped the old requirement to hold four meetings per financial year; two, or an equivalent level of activity, is now the floor. Four is still permitted.

No duplication. The Committee “may ask for exceptionally strong cases to be made where there is geographical and subject overlap between different Scheme 3 applications, or between the Scheme 3 application and an existing research network”. If a Joint Research Group already covers your area in your region, expect to explain what is different about yours — or, better, talk to the existing group before applying.

How applications are assessed

The LMS publishes the checklist its Committee uses. For Scheme 3 the criteria are:

• mathematical quality of the project;
• a well-argued proposal, with a clear statement of the mathematical component where the project is broader;
• a well-thought-out and cost-effective budget, restricted to allowable costs, with an appropriate and clearly identified balance of funding from other sources;
• the difference the Society’s grant makes to the viability and success of the project;
• likely beneficiaries;
• compliance with the Society’s policy on Women in Mathematics and its expectations of those seeking grants;
• a limit on the number of grants awarded to any one individual or department;
• sufficient time available for publicity of the event;
• receipt of satisfactory reports on previous grants.

Several of these are worth reading twice. “The difference made by the Society’s grant” is a counterfactual question: what would not happen without £2,400? An honest answer — “without support, the Newcastle and Aberdeen members cannot afford to attend, and the group reverts to a Manchester–Sheffield seminar” — is far stronger than an assurance that the money would be helpful.

“Limitation on the number of grants awarded to any individual or department” means that if your department already holds two Scheme 3 grants, a third from the same place is a harder sell. Spread grant-holding around the nodes across successive years.

“Receipt of satisfactory reports on previous grants” is the quiet killer for renewals. Reports are due by 30 September each year — the same date as the application deadline. A renewal application submitted on time with the previous year’s academic and financial report outstanding is at real risk.

The diversity requirement is not decorative. The Society states elsewhere in its grant guidance that a failure to include women speakers or participants is grounds for refusal of funding. The LMS publishes a Good Practice sample answer for applicants to work from, covering access needs collected at registration, sharing the LMS Code of Conduct and naming two contact points for concerns, aiming for gender balance among speakers and panel members, scheduling that supports early-career researchers and those with caring responsibilities, and practical wellbeing and sustainability measures such as adequate breaks, vegetarian catering, reusable cutlery, minimal printing and public transport. Do not copy the sample verbatim; adapt it to what you will actually do, because you will be reporting against it.

How to apply

1. Confirm your three nodes and pick a grant-holder. The grant-holder carries the financial responsibility, so choose someone whose institution can administer reimbursements without friction.
2. Check the LMS member requirement. If the grant-holder is not a member, line up a current member to support the application before you start writing.
3. Draft the academic case. The LMS is specific about the angle: the justification “should focus on the benefit to UK mathematics that the proposed research meetings would bring”. Not the benefit to your subfield in the abstract, and not the benefit to the three applicants — the benefit to UK mathematics.
4. Build the budget. Travel, accommodation and subsistence within the £160 daily cap, caring costs, and any hybrid or online support. No honoraria. State clearly what other funding the group has, including any departmental contribution or registration income.
5. Complete the online form. Applications go through the LMS online application form, which requires a website account — register before deadline week rather than during it. Keep the application self-contained: the Society asks applicants not to append substantial documents full of irrelevant detail, and not to refer reviewers to a website for key information.
6. Get your two supporters to email in. The form must be endorsed by two supporters who confirm they have seen the application and agree to participate. That confirmation is emailed separately by the supporters to [email protected]. This is the single most commonly missed step, because it depends on other people acting after you have submitted.
7. Attach last year’s reports if you are renewing. Financial and academic report for the previous year, where applicable.
8. Submit ahead of the first meeting. The LMS asks that applications be submitted well in advance of the date of the first meeting, so do not plan an October event on the assumption of a decision that arrives the same month.

Timeline

Common mistakes

Treating the third institution as a formality. The most frequent structural weakness is a group that is really a two-department collaboration with a third name attached. The rotation of meeting venues, and the involvement of the third node in choosing speakers, are what the Committee is looking for.

Budgeting for honoraria. It is prohibited. If your discipline norm is to pay invited speakers, that has to come from somewhere else.

Assuming £2,400 will be awarded. The guidance notes state plainly that the published maximum is an upper limit, not a standard award. Build a programme that still works at a lower figure, and say in the application which meetings are the priority if a partial award is made.

Forgetting the supporter emails. The application is not complete until two supporters have written in. Ask them before you submit and give them the deadline.

Leaving the reports until the application. Reports and applications fall on the same date. Doing both in the last week of September, every year, is how groups end up with an unsatisfactory report on file.

Writing an academic case about the mathematics only. Mathematical quality is one criterion of nine. Beneficiaries, cost-effectiveness, counterfactual impact, publicity lead time and inclusive practice all carry weight, and they are the parts applicants most often skim.

Frequently asked questions

Can PhD students be involved? Yes — enthusiastically. They simply cannot be the applicant. Student travel and subsistence are legitimate costs, and the Society expects meetings to be advertised to research students and early postdocs.

Can we include a node outside the UK? Yes, provided at least two of the institutions are in the UK, Isle of Man or Channel Islands. The Society specifically encourages including nodes from countries where mathematics or mathematicians are under threat.

Can meetings be online or hybrid? Yes. The grant can contribute to the costs of running hybrid meetings, including caring costs for people attending remotely, and the LMS publishes separate guidance on supporting hybrid events.

Do we have to hold four meetings? No. That requirement was dropped. Two meetings, or an equivalent level of activity, is the minimum; four remains permitted.

Is this a multi-year grant? No. The award covers meetings over one year and must be claimed within the financial year it is awarded. Successful groups reapply annually, and satisfactory reporting is part of the assessment for the next round.

What if a similar group already exists in our area? Expect to make an exceptionally strong case where there is geographical or subject overlap with another Scheme 3 group or an existing research network. Contacting that group first is usually more productive than competing with it.

If Scheme 3 is not the right fit

The LMS runs a family of research grants with a common September–October deadline cluster, and the right scheme depends on the shape of the activity rather than its subject. Conference Grants (Scheme 1), Visits to the UK (Scheme 2), Research in Pairs (Scheme 4) and Collaborations with Developing Countries (Scheme 5) all close on 15 September 2026. Postgraduate Research Conference Grants (Scheme 8) and Celebrating New Appointments (Scheme 9) close on 15 October 2026. LMS Early Career Fellowships close on 1 October 2026.

For groups whose ambitions outgrow a few thousand pounds, the LMS itself points to the Isaac Newton Institute’s network support stream for the mathematical sciences, which makes awards of £5,000 to £25,000, is open to all, and has two deadlines a year — 31 January and 31 July. A sensible progression for a young community is to establish itself on Scheme 3, build a track record and an attendance list, and then apply to the INI for the larger sum.

Official links and contacts

• Scheme page and application form: https://www.lms.ac.uk/grants/joint-research-groups-uk-scheme-3
• LMS grant schemes guidance notes, including the Scheme 3 assessment checklist and the Good Practice sample answer: https://www.lms.ac.uk/grants/lms-grant-schemes-guidance-notes
• Calendar of LMS funding application deadlines: https://www.lms.ac.uk/grants/calendar
• Currently supported Joint Research Groups: https://www.lms.ac.uk/grants/joint-research-groups-uk-scheme-3/current
• Grants Administrator: Lucy Covington, [email protected]
• Chair of the Research Grants Committee: Professor Amanda Turner

The LMS states that queries about applications can go to either the Grants Administrator or the Committee Chair, and that both are willing to discuss proposals informally and advise on submission. For a scheme where the difference between funded and unfunded often comes down to whether the panel believes the third node is real, an informal conversation in August is worth more than another revision of the academic case in late September.