Trail Making Test
25 circles, scattered across a field. In part A you connect them in order, 1 to 25. In part B you alternate between two sequences — 1, A, 2, B — to the end. Both parts are timed, and the difference between the two times is the number this page reports. Free, no account, everything drawn and timed in your browser.
Take the test
Part A. Tap the circles in order: 1 to 25.
Part B. Alternate the two sequences: 1, A, 2, B … up to 13.
READY
Part A done. Part B puts two sequences on the field and asks you to alternate between them — 1, then A, then 2, then B, and so on. Nothing is being timed until you press the button.
0
seconds, part B minus part A
That gap is what the second sequence cost you. Part A already covers finding circles and getting your hand to them, so taking it away leaves the part where two sequences had to be kept going at once.
Part B came out the faster of the two here. It happens: a first part spent learning how the field is drawn can be the slower one, and a single pair of times is a noisy thing.
- Part A
- —
- Part B
- —
- Wrong taps
- —
Nothing is uploaded; both times stay in this tab. Tap a wrong circle and the clock keeps running.
That was a single run. In the app it becomes a short daily session, with a Focus Index to show for it.
Point your camera at the code — the App Store opens.
What the trail making test measures
Part A is a search task with a motor tail. The circles are numbered, you always know what comes next, and the whole of the difficulty is finding it on the field and getting your hand there.
Part B keeps every bit of that and lays a second sequence underneath it. Now each step also has to answer a question you did not have to ask in part A — was I on numbers or on letters? — and answer it right every time. Psychologists call that set shifting: holding two rules and moving between them without dragging one into the other. In an ordinary day it is what happens when a message arrives in the middle of something you were writing.
Neither part asks you to know anything. Counting and the alphabet are sequences you have had since childhood, and that is the reason the task is built out of them: with nothing to work out, what is left to be slow at is the searching and the switching. It is also why the task needs so little translating — there is no vocabulary in it and no arithmetic — though part B does still assume you have the Latin alphabet by heart.
Why the result is part B minus part A
Two times are interesting. The number between them is the one this page puts on top, and the reason anyone runs both parts rather than one.
Everything the two parts have in common sits in both times in roughly equal measure: how good your eyesight is, how big the screen is, how fast you move a hand, how much you care this afternoon. Subtract one time from the other and most of that cancels, and what survives the subtraction is the thing only part B asked for — running two sequences at once across 25 circles. That is why the card shows both times and puts the difference on top.
It is a subtraction, though, and not a purified measurement. Part B is longer to read as well as harder to hold, the letters sit on the field alongside numbers you have to keep ignoring, and any of that can land in the gap. Read the difference as the cost of the second sequence on this field, on this device, this afternoon — a quantity worth watching against your own earlier runs, and not a coefficient of anything.
Part A and part B, side by side
Same field, same clock, same circles. The only difference is what is printed on them and in what order they have to be joined.
Part A holds the numbers 1 to 25 and asks for counting order. Part B splits the same 25 circles between two sequences — the numbers 1 to 13 and the letters A to L, 12 of them — and asks you to take one from each in turn: 1, A, 2, B, and so on. The alternation opens on a number and closes on one, which is why there is one more number than there are letters.
A fresh field is drawn for each part, and again for every run. If the circles stayed where they were, your second attempt would measure how well you remembered the board rather than how fast you can read a new one — and part B on part A’s field would measure that twice over.
Where it comes from
- 1944
An army sorting sheet
The task appears in the Army Individual Test Battery, issued by the US War Department: two sheets of circles, a pencil and a stopwatch, used to sort recruits. Nothing about it was designed for a clinic — it was cheap, quick, and hard to talk your way through.
- 1950s
Reitan takes it into neuropsychology
Ralph Reitan publishes on the test in the mid-1950s and takes it into the Halstead-Reitan battery, where it has stayed ever since. That is how an army sheet became one of the most administered tasks in the field — and why nearly everything published about it describes a pencil, a sheet of paper and a person holding a stopwatch.
- 2004
Norms — for the paper version
Tom Tombaugh publishes normative data for both parts, stratified by age and by years of education, in Archives of Clinical Neuropsychology. It is the table most often quoted for this test, and every number in it was collected on paper under standard administration. That is the whole reason no figure from it appears on this page.
- Today
The same task, different hardware
The format moved onto screens, which is why you can take it here. What did not move with it is comparability: a pencil dragged between circles and a finger tapping them are different acts, and the times they produce sit on separate scales that only compare to themselves.
How to read your result
Three things on the card, and only one of them is the headline.
-
The two times
Seconds from Start to the last circle, once for each part. Both are printed, because a difference with its two terms hidden is a number nobody can check.
-
Part B minus part A
The headline. What the second sequence cost you once the searching and the moving have been subtracted out — the reason both parts are run rather than just the harder one.
-
Wrong taps
Circles hit out of turn. The clock keeps running and the target does not move on, so an error has already cost you time; the count says where it went. A quick part B bought with a handful of them is not a quick part B.
What this version cannot tell you
-
A mouse is not a pencil
On paper the line is dragged from circle to circle and the pencil never leaves the page. Here you tap, or you click. Those are different motor tasks with different costs, and both of them are inside your times — as is the size of the screen you are holding.
-
One run is mostly noise
Both times swing with how alert you are, how much light is in the room and where the circles happened to land. Two runs an hour apart can differ by a good deal with nothing about you having changed.
-
You improve at it quickly
Most people get noticeably quicker over their first few attempts. That is learning how the field is drawn and where to look on it, not a change in anything the test is named after. Expect the gains, and compare runs once they have flattened.
-
It is not a diagnosis
The trail making test is used in clinical assessment, but never alone: it sits inside a battery, is given under controlled conditions and is read by someone qualified alongside a history and an interview. One task in a browser tab is not that. A slow run here means you had a slow run here — and if your own thinking, or that of someone close to you, genuinely worries you, that is a conversation for a clinician rather than for a web page.
Questions about the trail making test
What is the trail making test?
A timed connect-the-circles task in two parts. Part A scatters the numbers 1 to 25 across a field and asks you to join them in order as fast as you can. Part B puts the numbers 1 to 13 and the letters A to L on the same kind of field and asks you to alternate — 1, A, 2, B — to the end. Both parts are timed, and the difference between the two times is the figure this page reports. It came out of an army test battery in 1944, and it is now taken in browser versions as well as on paper.
What is the difference between part A and part B?
Part A of the trail making test runs through one sequence you have known since childhood, so the only hard part is finding the next circle. Part B runs two sequences at once and makes you take one from each in turn, so on top of the searching you have to keep track of which sequence you are on and not carry the other one into it. Everything else is identical, which is the point: what part B costs you above part A is what running two sequences costs, on this field, on 25 circles.
What is a good trail making test time?
We are not going to name a good trail making time, and the reason is not modesty. The figures in circulation come from the paper test — a printed sheet, a pencil, an examiner with a stopwatch — and a mouse or a thumb on a screen is a different instrument that produces different seconds. Quoting one beside the other would look authoritative and would be wrong. What can be said honestly: part B takes longer than part A for very nearly everybody, and the comparison worth making is with your own earlier runs, on the same device.
Are there age norms for the trail making test?
Published tables for the trail making test exist, stratified by age and by years of education, and they are genuinely useful to the people who administer the paper test under the conditions those tables were built on. None of that transfers to a browser: the sheet is not the same sheet, the pencil is a cursor, and nobody is timing you with a stopwatch across a table. So this page prints no age table and describes no time as normal or abnormal. Anywhere online that does is either using a procedure you cannot check or quoting numbers it has no right to.
Why is part B so much slower than part A?
Because part B of the trail making test asks for two things at once. Every step in part B is a search — the same search part A asks for — plus a decision about which of two sequences you owe next, and both of those cost time. Some of the gap is also plain reading: two kinds of symbol on one field take longer to tell apart than one kind does. That is exactly why the result is a difference rather than part B alone; subtracting part A takes the searching out and leaves most of the switching behind.
What happens if I tap the wrong circle?
On the trail making test a wrong circle counts as a wrong tap, the clock keeps running and the target does not move on — you still have to find the circle you were looking for. That follows the logic of the paper procedure, where the examiner sends the pencil back to the last correct circle and does not stop the stopwatch to do it. The count is shown next to your times rather than folded into them, because a fast run full of wrong taps is a different thing from a fast clean one.
Is this the same test they use in clinics?
No — this page is not the clinical trail making test. The task is the same idea, and the procedure is not: our fields are generated in your browser rather than printed from a standard form, the circles are a different size in a different arrangement, and you tap instead of drawing a line. It is also not administered — nobody is watching you, explaining the rules or noticing that you misheard them. Treat this as a version of the task you can try, not as the instrument a clinician would use, and never as anything that tells you about your health.
Do I need an account?
No. The trail making fields are generated in your browser and both times are measured there — nothing is uploaded, nothing is stored, and your result is gone when you close the tab. There is no email box anywhere on this page.
One afternoon is a number. A month of them is a picture.
A gap measured once moves with how alert you happened to be. In NEURA the tasks adapt to you, run inside a short daily session, and roll up with the rest of the catalog — 19 games — into a single Focus Index you can watch over weeks. That is the part one page, played once, can never show you.
