Your Neurologist Asked for a Headache Diary. Record This
Six fields, and what a clinician does with each one at the next appointment — plus the four things people record diligently for months that never get used.

They asked for a diary, said “just keep track of them”, and moved on to the prescription. Nobody said what to write down.
So most people either record nothing for eleven weeks and reconstruct it in the car park, or record everything and abandon it by week three. The version that survives is six fields, and the reason it is those six is that each one feeds a decision the clinician is going to make in front of you.
The six fields, and what each one is for
The date. Not the week, the date. Days are the unit every threshold in headache medicine is written in, and an attack running from Tuesday evening into Wednesday afternoon is two days rather than one event. That distinction decides which side of a line you fall on, which how many migraines a month counts as chronic works through with a full month counted both ways.
What it feeds: the migraine-days and headache-days figures, which route almost everything after them.
Was it migrainous, or an ordinary headache? A yes, a no, or an honest “not sure”. Both numbers get asked for separately, and people who only count the bad ones understate their total by a third.
What it feeds: whether your pattern meets a definition, and whether an increase is in migraine days or in background headache — two different problems.
Roughly how long it lasted. To the nearest few hours is fine. Record when the pain stopped, separately from the foggy day afterwards.
What it feeds: whether a drug is covering the length of your attacks. A short-acting tablet against a two-day attack is a recurrence problem rather than a failure, which is a different conversation and a different fix.
What you took, and at what time. The name and the clock time. The gap between onset and dose is the single most decision-relevant number in the whole diary, and it is the one nobody can reconstruct by the following morning.
What it feeds: two things at once — your monthly medication-day count, and whether apparent drug failures are actually timing failures.
What it stopped you doing. One short line. “Left work at two.” “Lay in the dark, couldn’t follow a conversation.” “Worked through it but not in the open-plan office.”
What it feeds: the disability picture, which is what treatment intensity gets matched to. It is also far more legible than a number out of ten, for reasons how to describe migraine pain to a doctor goes into.
A tick on the clear days. Ten seconds. No attack, and roughly how you slept.
What it feeds: the denominator. Without it you have a list of bad days and no way to express a rate, and a rate is the only thing that can show a preventive worked.
That last one is the field people cut first and the one that quietly makes the rest usable.
What gets recorded for months and never used
Long descriptions of the pain. Two years of vivid free text about what each attack felt like. It reads as diligence and functions as noise: the diagnostic features are a short fixed list, and once they are established, the adjectives add nothing. Write the quality once, at the start, and stop.
A parallel food diary. Enormous effort, and it rarely survives contact with the fact that food cravings are a common premonitory symptom — the attack was often already underway when the chocolate started looking appealing. How to track migraine triggers without fooling yourself covers why this particular record produces confident wrong answers.
Weather written down from memory. Recorded the following day, from what you remember of the sky, this measures your beliefs about weather rather than the weather. It is also the one variable memory cannot supply at all.
A severity scale that has drifted. A 7 in January becomes a 5 by June once you have had worse, so the trend line slopes downward while nothing has improved. Fix the anchors in writing before you start — what your 5 means, what your 9 means — and check back against them every few months.
Bring the summary, not the diary
The most common way a well-kept diary fails is at the moment of handover.
Ninety pages of daily entries cannot be read inside a ten-minute appointment. They get set aside, and the time you needed for the actual questions goes on flicking through them. What works is one page: the four figures, a calendar of the last three months with attack days marked, and your treatment history. What to bring to a neurologist appointment sets that page out in full.
Keep the detail in reserve. If they want to dig into a particular fortnight, you have it.
Three months, marked on the day
One month is the usual minimum. Three is what the definitions are written against, because they describe patterns sustained beyond three months rather than a single rough stretch.
The discipline that matters more than any field: mark it the same day. By the next morning the start time has gone, the severity has softened, and the previous day has quietly reordered itself into a story that makes sense — usually the story you already suspected. Retro-filling produces a tidy record of your beliefs.
Paper is genuinely good at this. The free printable diary is a monthly grid of exactly these six columns, with a separate sheet for attack detail so a bad day never asks you to fill in twelve fields. Migraine Journal records the same six in under a minute and stamps the 24-hour barometric-pressure change automatically, which is the only way that particular column ever gets filled in honestly.
Whichever you use, the diary is a record rather than a diagnosis, and what it appears to show belongs in the appointment rather than in a decision you make alone. The rest of the appointment guides cover what happens once it is in front of someone.
Quick answers
What should you record in a headache diary for a neurologist?
Six fields: the date, whether it was migrainous, roughly how long it lasted, what you took and at what time, what it stopped you doing, and a tick on the clear days. Everything else is optional and most of it never gets used.
Why record the days with no headache?
Because without them a diary is a list rather than a rate. Treatment decisions are made against days per month, and a record that only contains bad days cannot produce that figure — it can only prove that bad days happened.
How long should you keep it before the appointment?
Three months if you have them, one month at minimum. The definitions a clinician works from describe patterns sustained over three months, and shorter runs are where people draw confident wrong conclusions from an unusually good or bad stretch.
What gets recorded but never used?
Long free-text descriptions of the pain, food diaries kept alongside, weather noted from memory, and a severity scale that quietly drifts over months. All four cost real effort and none of them changes a decision.
Walk in with the numbers, not an impression
Attack count, average severity, acute-medication days, and a printable summary of the last 30 days — the four things the questions in this post are really asking for. MIDAS and HIT-6 are built in, so the disability score is already calculated when you arrive.
This article is general information, not medical advice. Talk to your GP or neurologist about your own symptoms, medication, and treatment.


