Exam scheduling

Exam timetable generator

An exam timetable runs once. Every mistake in it is discovered by a student sitting in the wrong hall — which is why it needs different tooling from the weekly grid.

Start free trial See the live demo

A weekly timetable that contains an error gets corrected in week one and runs correctly for the remaining thirty-eight. An exam timetable that contains an error is discovered live, by a candidate sitting the wrong paper at the wrong time — and there is no second attempt. The two problems look similar and have completely different tolerances.

That difference is why exam scheduling defeats tools built for the weekly grid, and why it is worth understanding what actually makes it hard before choosing anything.

Exam clashes are student-level, not class-level

In a weekly timetable a class is a unit. Year 9 Set 2 moves as a block, so checking the block is enough. Exams do not work that way. Students sit individual option combinations, so a clash between Geography and Computer Science on Thursday morning may affect four candidates in the entire cohort.

No amount of checking at group level will find that. You have to test every student’s personal entry list against every session — which is exactly the kind of exhaustive cross-referencing people are bad at and software is good at.

Half your constraints are set by someone else

Awarding bodies fix the dates and times of external papers. Your scheduling freedom exists only in what they leave behind: which room, which invigilators, which seating, and what to do about candidates whose externally-fixed papers collide.

That last one has a long lead time — resolving a genuine clash usually means a formal arrangement with the board. Finding it in March is a process. Finding it in May is a problem.

Access arrangements change the room and the staffing

Extra time means a candidate finishes after the room booking ends. A separate room means an invigilator you may not have counted. A reader or scribe is a person who must be scheduled, qualified, and not simultaneously invigilating elsewhere.

These are not preferences to balance. Getting them wrong is a compliance failure with consequences for the student and for the centre — and they are frequently tracked on a spreadsheet running alongside whatever built the timetable, which is where the two versions start to disagree.

Invigilation is a second scheduling problem inside the first

Invigilators must cover a full session, cannot supervise a subject where a conflict of interest exists, need breaks across a long day, and are usually a mix of staff and externally engaged people with limited availability.

Schools routinely build the exam timetable first and discover the invigilation shortfall afterwards — at which point the timetable is fixed and the shortfall is something to absorb rather than avoid. Treating invigilation as a constraint on the plan, not a task after it, is the single change that prevents this.

Exam-condition capacity is not teaching capacity

A hall that seats 200 for assembly might seat 90 under exam spacing rules. Planning against the teaching number is a mistake discovered at the worst possible moment, with candidates already in the building.

The order that works

Lock the externally fixed dates first — they are immovable
Find student-level clashes across the whole cohort before touching rooms
Apply access arrangements, which change room and staffing needs
Allocate rooms against exam-condition capacity, not teaching capacity
Schedule invigilation as a constraint, not an afterthought
Publish personal schedules to candidates, not a master grid on a wall

Frequently asked questions

Can I build an exam timetable with the free demo on this site?

No. The demo on this site schedules a weekly class timetable for a single small year group, to show how the constraint engine works. Exam scheduling needs student-level entry data, access arrangements and invigilator availability, which is Skoolia territory.

What is the most common exam timetabling mistake?

Checking for clashes at group level rather than per student. A clash affecting four candidates out of two hundred is invisible in the master grid and obvious to the four students it affects.

Should exam scheduling use the same system as the academic timetable?

It should use the same data. Student option choices, access arrangements, staff availability and room capacities already exist in the school’s records; re-entering them into a standalone exam spreadsheet duplicates the work and creates a second version that can disagree with the first. The exam version is the expensive one to get wrong.

How far ahead should exam timetabling start?

Early enough that a genuine clash can still be resolved with the awarding body, which means as soon as entries are confirmed. The scheduling itself is quick; the lead time is for the exceptions.

Generate your school timetable with AI

Start a 14-day free trial on Skoolia — no credit card, cancel anytime.

Start free trial

14 days free · No card required · Clash-free timetables in minutes