Stop-Signal Task online: SSRT and response inhibition

Arrows point left or right and you press the matching arrow key, fast. On 1 trial in 4 a red square follows the arrow and you try not to press. The result estimates how long stopping takes you.

Start the Stop-Signal test

200 trials · about 6 min · arrow keys

After the last trial you see your stopping time and can keep it with a free account.

200 trials · about 6 min · arrow keys

This task needs a keyboard with arrow keys and a larger screen. Open this page on a computer to take it.

Start the Go / No-Go test

Go / No-Go works on a phone: tap for every letter except X.

SSD
The arrow, then the squareA stop trial · the gap is the SSD

How it works

16 practice trials with feedback come first, then 200 scored trials in one run without feedback. Most trials last a second or two, so the whole task takes about 6 minutes.

  1. 1 · Practice

    16 trials with feedback

    12 arrows and 4 stop trials in random order. After each one, a tick, a wrong-direction note or a reminder to stop on the square shows how it went. Practice is not scored.

  2. 2 · 150 trials

    Press the arrow

    An arrow appears and you press the arrow key that matches it, as fast as you can. You have 1.5 seconds, then the screen stays blank for half a second.

  3. 3 · 50 trials

    Hold back on the square

    The arrow appears as usual, and after a short delay a red square joins it. Try not to press. A press before the square shows, or after, counts as a failed stop.

  4. 4 · Stop trials

    The delay adapts

    The delay starts at 250 ms. Each time you stop, the next one is 50 ms longer and harder; each time you press, it is 50 ms shorter. It stays between 50 and 800 ms.

Go and stop trials are shuffled, so you never know which kind is next. The staircase is meant to bring you close to stopping half the time, which is why the task feels hard and why missing stops is expected. A thin bar along the top of the screen tracks your progress.

The stop-signal task is a standard research tool for timing how quickly a response can be cancelled. In 2019 a group of researchers who use it published 12 consensus recommendations for running and scoring it, because design and analysis choices that vary between studies can easily change the result (Verbruggen et al., 2019). This version is a choice task between 2 arrows with a staircase on the delay.

What it measures

Every trial starts the same way, with a quick press already on its way. The stop trials ask whether you can cancel it in time.

Go trials

  • Choice reaction time: how quickly you press the key that matches the arrow.
  • Accuracy: pressing the right direction and not letting arrows go by.
  • Keeping up a steady pace over 150 go trials with no feedback.

Stop trials

  • Response inhibition: cancelling a press that has already started.
  • Stopping speed, estimated as SSRT, the stop-signal reaction time.
  • Not giving in to the pull to wait for the square, which would slow every go press.

SSRT cannot be timed directly, because a successful stop leaves no press to time. The task estimates it. Picture each stop trial as a race between your press and your stopping. When the delay settles where you stop about half the time, stopping finishes, on average, about when your usual press would have landed. Take the delay away from your usual go time, and what remains estimates how long stopping took.

Stopping has been studied in many groups. Meta-analyses have compared stopping performance between people with and without ADHD (Lijffijt et al., 2005), but a task score does not identify anyone. In a large online study of self-control tasks and questionnaires, the tasks, this one among them, showed little relationship with what people reported about their own self-control (Eisenberg et al., 2019).

A test, not a diagnosis.

An SSRT comes from one session on one task. On its own it cannot confirm or rule out any condition.

What you get

The result screen opens as soon as the 200th trial is over.

On the result screen

  • Your SSRT, your mean go reaction time and your mean stop-signal delay, in milliseconds.
  • A summary line and a marker on a fast-to-slow scale, both set by fixed SSRT bands: under 200 ms, 200 to 250, 250 to 300 and over 300. The bands are the same at every age.
  • Notes on your go accuracy and on how often you stopped, which say how far the estimate can be trusted.
  • A data panel with every figure and the trial counts, and a CSV file of the scored trials.

When you keep it

  • Saved to your results when you create a free account, and there on any device you sign in from.
  • Listed under attention, next to Go / No-Go, Flanker and the CPT once you take them.
  • A next test is suggested as soon as you finish.

How this version is scored

The test records every trial: the arrow, the delay, your key and its timing. Each figure comes from that record, with practice left out.

Go reaction time
The mean time from arrow to key press, over the go trials where you pressed the right key within 1.5 seconds. Wrong keys and missed arrows are left out of the mean.
Go accuracy
Correct presses as a share of all 150 go trials. A missed arrow lowers it as much as a wrong key.
Stop rate
The share of the 50 stop trials with no press at all. The staircase aims to hold it near 50%.
Mean SSD
The average stop-signal delay over the last 25 stop trials, the second half, once the staircase has had time to settle.
SSRT
SSRT = mean go reaction time − mean SSD. Shorter means faster stopping. This is the mean method, the simplest way to estimate SSRT from a staircase.
No SSRT shown
The estimate appears only when at least 75 of the 150 go trials were correct, you stopped on 25% to 75% of stop trials and the result is above 0. Otherwise the screen says there were not enough valid responses, shows your raw counts and suggests a retake.

Why waiting does not pay. Slowing down to check for a square makes stopping easier, so the staircase answers with longer delays. Your go time rises too, and the subtraction takes much of that gain back out. The estimate works best when you press as fast as you can and accept the failed stops.

Other ways to estimate SSRT. Researchers also estimate SSRT with methods that use the whole spread of go times instead of the mean, and these can give somewhat different numbers from the same trials. Compare an SSRT only with figures from the same method and design.

Before you start

The estimate rests on fast, steady key presses, so set up before you begin.

For a cleaner attempt

  • Use a computer with a physical keyboard and rest two fingers on the left and right arrow keys.
  • Sit somewhere quiet and silence notifications for the next few minutes.
  • Press as soon as the arrow appears. Do not hold off to see whether a square follows.
  • Expect to miss about half of the stops. The task is built that way.

Interpret cautiously when

  • You were tired, unwell, rushed or interrupted. Short-term sleep loss affects attention tasks (Lim and Dinges, 2010).
  • Your keyboard, an assistive device or a slow computer delays key presses.
  • Your go accuracy was low or your stop rate ended far from 50%, which the result screen points out.
  • You have taken the task before; knowing its rhythm can change how you approach it.

Questions

What does the stop-signal task measure?

It measures response inhibition: how quickly you can cancel a response that has already started. The main score is SSRT, the stop-signal reaction time, which the task estimates from your go reaction time and the delay at which you stop about half the time. The go trials also give a reaction time and an accuracy.

What is a normal SSRT?

It depends on the version and the scoring method, so there is no single cut-off. In 72 healthy adults aged 40 to 77, tested with a staircase like this one, the mean SSRT was about 302 ms (Lee and Hsieh, 2017), and the open data of a large online study put adults aged 18 to 45 near 286 ms. The result screen reads your SSRT against fixed bands that place 200 to 250 ms in the middle, below both of those averages, so treat the bands as a rough guide.

Can I take the Stop-Signal test on my phone?

No. The tests hub asks for a computer with arrow keys, because the score rests on the exact timing of each key press. On a phone you can start with Go / No-Go, which also asks you to hold back a press, and come back to this one on a computer.

Why can I not stop on every red square?

Because the task is built so you cannot. Each successful stop makes the next delay 50 ms longer, so the square arrives closer to your press. The staircase keeps adjusting until you stop on roughly half of the stop trials, the level SSRT needs. Failed stops are part of the measurement, not mistakes to avoid at the cost of speed.

Sources

By Cristian Dominguez Rein-Loring, Creative Technology Lead · Updated October 12, 2026