Speed Test
Reaction Time Test
The classic reflex test: click the moment the screen turns green. We average five attempts for your score. Global average: 273ms.
Click to Begin
Wait for green, then click as fast as you can
Your Results
Don't lose this result
Create a free account to save every score and track your improvement.
What reaction time measures
Simple visual reaction time (SRT) — what most people call a reflex test — is the elapsed time between a stimulus appearing on screen and your motor response completing. It is the most fundamental benchmark of neural processing speed - Capturing the entire perceptual-motor chain from photon hitting your retina to your finger registering a click. Franciscus Donders first measured it systematically in 1868; over 150 years of research have established the global average at 250–280ms in laboratory conditions, and slightly higher in large web-based test datasets, where display and input latency are included in the measurement.
SRT is distinct from more complex variants. Choice reaction time adds a stimulus-identification decision that costs ~100–150ms. Audio reaction time uses sound instead of light and is typically 30–50ms faster. Aim Trainer adds a motor targeting component on top. This test isolates the pure detection-to-response interval - The irreducible floor of human response speed.
The neural reaction chain
The theoretical minimum for genuine human reaction (not anticipation) is approximately 100–120ms - The combined minimum neural conduction time. Scores below 100ms are anticipation clicks. The remaining gap between the 120ms floor and your actual score represents variance in attentional readiness, arousal state, and motor preparation speed.
Reaction Time Score Distribution
The distribution is right-skewed: most scores cluster between 220–320ms, with a sharp lower tail (genuine fast responders) and a long upper tail (fatigue, mobile devices, and hardware lag). Trimmed means that exclude the slowest tail typically land in the 260s; the raw ~273ms web average is inflated by that upper tail.
Reaction time distribution (ms)
Score percentile reference
| Percentile | Reaction time | Classification |
|---|---|---|
| Top 1% | <160ms | Exceptional |
| Top 10% | 160–210ms | Excellent |
| Top 25% | 210–250ms | Above average |
| 50th (median) | 250–300ms | Average |
| Bottom 25% | 300–360ms | Below average |
| Bottom 10% | >360ms | Slow |
Benchmarks based on published norms and large public datasets, trimmed to exclude likely anticipation clicks (<100ms) and extreme hardware lag (>800ms). Compare your score on the reaction time leaderboard.
Visual vs audio vs tactile reaction time
Reaction time is not a single number - It varies by sensory modality. The auditory pathway to the motor cortex is shorter and has fewer synaptic relays than the visual pathway, making audio reaction time consistently faster. Tactile (touch) RT is faster still but rarely tested outside laboratory settings.
| Modality | Average RT | Why faster/slower |
|---|---|---|
| Tactile (touch) | ~155ms | Shortest pathway - Skin receptors directly activate spinal reflexes |
| Auditory (sound) | ~170–200ms | Cochlea → auditory cortex has ~4 synaptic relays vs ~6 for vision |
| Visual (this test) | ~250–280ms | Retina → V1 → motor cortex adds phototransduction delay (~20ms) |
Compare your visual and audio RT on the same day to estimate your personal hardware offset. If the gap is more than 70ms (rather than the expected 30–50ms), Bluetooth audio latency is likely inflating your audio score. The Reaction Time FAQ covers hardware calibration in detail.
Hardware latency: the hidden variable
Your measured RT is always neural RT + hardware RT. On a well-configured desktop, hardware adds 6–20ms. On a 60Hz monitor with a wireless mouse and Bluetooth audio, hardware can add 50–150ms - Completely swamping genuine individual differences. This is why direct comparisons between users on different hardware are imprecise.
| Hardware | Added latency | Notes |
|---|---|---|
| Wired mouse | 0–2ms | Negligible. Recommended for testing. |
| Wireless mouse (2.4GHz) | 1–8ms | Modern low-latency wireless is close to wired. |
| Bluetooth mouse | 8–30ms | Variable. Avoid for precise measurement. |
| 240Hz monitor | ~4ms | Fastest common display update rate. |
| 144Hz monitor | ~7ms | Good. Small difference from 240Hz in practice. |
| 60Hz monitor | ~17ms avg / 33ms max | Largest source of display lag. Most laptop screens. |
| Phone touchscreen | +30–80ms | Touch registration latency. Highly variable by device. |
Best practice for accurate measurement: Use a wired mouse, a 144Hz+ monitor, and close all other browser tabs. Three consecutive runs on the same hardware give a representative baseline. Hardware-corrected scores for Aim Trainer and Choice Reaction Time should be taken on identical hardware for valid comparisons.
Factors that affect your score
Reaction time fluctuates by 15–40ms across sessions in the same individual due to transient state factors - Arousal, hydration, fatigue, and caffeine. Understanding these lets you control for them when comparing across sessions, and explains why a "bad day" score is often 30–50ms slower than your true resting baseline.
| Factor | Effect on RT | Evidence |
|---|---|---|
| Sleep deprivation (2hrs short) | +20–40ms slower | Strong |
| Caffeine (100–200mg) | −10–30ms faster | Moderate |
| 3-week deliberate practice | −10–20ms faster | Strong |
| Phone touchscreen vs wired mouse | +30–100ms slower | Strong |
| Age (per decade after 25) | +2–5ms per decade | Strong |
| Aerobic exercise (30min prior) | −8–15ms for 1–2h | Moderate |
| Action video games (20–50hrs) | −15–25ms faster | Moderate |
| Alcohol (0.08% BAC) | +30–50ms slower | Strong |
| Time of day (morning vs afternoon) | +10–20ms at peak hours | Weak |
A single night of poor sleep (5–6hrs) slows RT by 30–50ms on average. Total sleep deprivation for 17–19 hours produces impairment equivalent to 0.05% blood alcohol - Legal intoxication in many countries. We cover how sleep quality affects reaction time in detail. See brain health FAQ for the sleep-cognition evidence.
RT improves with practice not because neural conduction velocity changes - It doesn't - But because attentional readiness increases, cutting the pre-processing latency before the motor command fires. Most gains occur within the first 50–100 trials. See reaction time FAQ for training protocols.
How to improve your reaction time
Genuine neural reaction time is constrained by biology - Myelination, synaptic transmission speed, and motor unit firing rate. However, the measured RT includes substantial modifiable components: attentional readiness, hardware setup, arousal state, and warm-up effects. Here is what the evidence supports:
Warm up with 3–5 practice clicks first
The first 2–3 clicks in any session are 15–30ms slower due to cold motor system activation. Always complete practice rounds before a scored attempt. This is the single highest-ROI improvement with zero cost.
Optimise hardware - Use a wired mouse on a 144Hz+ screen
Upgrading from a 60Hz screen to 144Hz cuts display lag from ~17ms to ~7ms. Using a wired mouse instead of wireless removes 1–30ms of variable input lag. Together these can improve your measured RT by 10–40ms with no change to your biology.
Moderate caffeine - 100–200mg, 45 minutes before testing
Well-documented 10–30ms improvement in simple RT at this dose. Effect peaks 45–90 minutes after ingestion. Tolerance builds with regular use, so habitual coffee drinkers see smaller acute effects.
Daily practice - 10 minutes over 3 weeks
Consistent practice reduces RT by 10–20ms through attentional readiness improvements. Training in Aim Trainer additionally builds the motor targeting layer. Action video games (20–50hrs of practice) show 15–25ms improvement in controlled studies.
Sleep 7–8 hours - The largest acute lever
Sleep deprivation degrades RT more than almost any other cognitive measure. Going from 6hrs to 8hrs of sleep reliably improves RT by 20–40ms. No other intervention comes close to adequate sleep for RT improvement. See brain health FAQ for the full evidence.
Reaction time by age
Simple visual RT peaks in the early-to-mid 20s at approximately 218ms (our platform median for age 20–24) and slows by roughly 2–5ms per decade thereafter. This decline reflects reduced myelination efficiency and slower central processing - Part of the same trajectory seen in processing speed decline. By age 65+, the average on this platform rises to ~310ms. Trained individuals - Athletes, musicians, action gamers - Maintain faster RT into their 40s and 50s compared to sedentary peers, but the underlying biological trend is not reversed. For a full breakdown of average reaction time by age, gender, and device, see our data deep-dive.
Age-related decline is further explored in the cognitive decline FAQ.
Clinical and real-world applications
Reaction time is one of the oldest and most widely used biomarkers in clinical and applied psychology. Its simplicity - A single number, one stimulus, one response - Makes it a reliable index across very different contexts.
Concussion screening
Simple RT is a sensitive marker for concussion - Even mild TBI slows RT by 20–50ms for weeks. Return-to-play protocols in contact sports use RT baselines to determine when athletes are fully recovered.
Parkinson's disease monitoring
RT slows progressively in Parkinson's, reflecting basal ganglia dopamine depletion. Longitudinal RT tracking is used to monitor disease progression and medication efficacy in research settings.
Driving safety
At 100km/h, 100ms of RT difference translates to ~2.8m of stopping distance difference. Regulators use RT data to establish safe driving limits for age, alcohol, and medication side effects.
Esports performance
Professional FPS players average 180–220ms on this test - Faster than general population but not dramatically so. Elite gaming performance relies more on motor targeting speed and decision-making than raw RT.
For a broader view of how reaction time fits into overall cognitive health, see the cognitive tests FAQ and explore complementary tests including processing speed, attention, and response inhibition.
Benchmark Numbers
On this page
Related tests
Site Record
Track your improvement
Create a free account to save your scores over time and see how you improve.
Sign up free