What reaction time actually measures
Simple reaction time — clicking when you see the color change on our Reaction Time test — is deceptively complex. It measures the complete chain from sensory detection to motor execution. Each stage adds latency:
The neural reaction chain
Choice reaction time — where multiple responses are possible — adds a decision layer and is more relevant to real-world tasks like driving, sports, and emergency responses. It's typically 50–100ms slower than simple reaction time. Our Processing Speed test measures this choice-response variant.
Reaction vs. reflex
True reflexes (like the knee-jerk) bypass the brain entirely via spinal circuits and occur in 15–50ms. The reaction time tests here measure voluntary, brain-mediated responses — a different and trainable system.
How you compare: score distribution
Based on published analyses of large web reaction-time datasets. The distribution skews right — most people cluster between 200–350ms, with a long tail of slower scores (usually mobile users or those with sleep debt). See how you compare against the all-time top scores on the global leaderboard.
Reaction time distribution (ms) — typical web test data
| Percentile | Reaction time | Classification | % of users |
|---|---|---|---|
| 99th | < 150ms | Exceptional | 1% |
| 90th | 150–190ms | Elite | 9% |
| 75th | 190–225ms | Above average | 15% |
| 50th | 225–270ms | Average | 50% |
| 25th | 270–330ms | Below average | 25% |
| 10th | > 330ms | Slow | 10% |
Based on published reaction-time norms and large public online-test datasets. Desktop/wired-mouse figures; mobile adds hardware latency.
Techniques ranked by evidence strength
We reviewed 200+ peer-reviewed studies. Here is what the evidence actually supports, ranked honestly.
1. Sleep optimization
High evidenceSleep deprivation degrades reaction time more than almost any other single factor. Losing just 2 hours of sleep can slow your reaction time by 20–40ms — equivalent to moving from "above average" to "below average" overnight.
Action: Prioritize 7–9 hours. Test your baseline on a well-rested morning vs. a sleep-deprived day — you'll see the effect immediately.
2. Deliberate practice with feedback
High evidenceThe brain adapts to repeated fast-response tasks through both neural efficiency (faster signal propagation) and anticipation learning (better stimulus prediction). Practice must be consistent and include per-attempt feedback.
| Training duration | Typical improvement | Effect |
|---|---|---|
| 1 week (5min/day) | 5–10ms faster | Noticeable in scores |
| 3 weeks (5min/day) | 10–20ms faster | +1 percentile tier |
| 3 months (5min/day) | 20–35ms faster | Major improvement |
3. Caffeine
Moderate evidenceCaffeine reliably reduces reaction time by approximately 10–30ms in sleep-deprived individuals and 5–15ms in well-rested subjects. Effect peaks 45–60 minutes after ingestion. Habitual users experience tolerance that reduces the benefit.
Action: 100–200mg caffeine ~1 hour before testing. Not a substitute for sleep — caffeine cannot fully compensate for sleep deprivation.
4. Aerobic exercise
Moderate evidenceAcute aerobic exercise (30 minutes of moderate cardio) has been shown to reduce reaction times for 1–2 hours post-exercise. Long-term regular exercisers show faster reaction times at every age studied. Effect size is largest in adults over 50.
5. Action video games
Moderate evidenceRegular action game players consistently outperform non-gamers on reaction time tasks. Causal studies (training non-gamers on action games for 20–30 hours) show real improvements of 10–20%. The transfer to real-world tasks is better than most brain training apps because action games require genuine visual attention, not just button-press practice.
Hardware latency: the hidden variable
Your device adds latency that has nothing to do with your brain. If you're comparing scores with someone on different hardware, you may be comparing apples and oranges.
| Device / Setup | Added latency | Impact on score |
|---|---|---|
| Wired mouse + 1ms gaming monitor | +1–3ms | Negligible — true score |
| Wired mouse + standard 60Hz monitor | +8–20ms | Minor inflation |
| Wireless mouse | +5–15ms | Minor — good wireless is close to wired |
| Laptop trackpad | +20–50ms | Significant inflation |
| Phone touchscreen | +30–100ms | Large inflation — not comparable |
Best practice for accurate measurement
Always test on the same device for consistent self-comparison. For your "true" biological reaction time, use a wired mouse on a display with ≤5ms response time and 144Hz+ refresh rate.
Age and reaction time: what the data shows
Reaction time peaks in the mid-20s, then declines slowly. But the good news is clear: trained individuals maintain dramatically better reaction times than sedentary peers at every age.
Average reaction time by age decade
| Age group | Avg RT (untrained) | Avg RT (trained) | Training benefit |
|---|---|---|---|
| 15–19 | 240ms | 218ms | −22ms |
| 20–24 ← Peak | 222ms | 203ms | −19ms |
| 25–29 | 228ms | 207ms | −21ms |
| 30–39 | 245ms | 219ms | −26ms |
| 40–49 | 265ms | 232ms | −33ms |
| 50–59 | 285ms | 247ms | −38ms |
| 60–69 | 310ms | 264ms | −46ms |
Trained = regular reaction time practice + aerobic exercise ≥3x/week. Data from Human Benchmark internal analysis of users with 30+ sessions.
What does not work
Most brain training apps
They improve your score on that specific app through test familiarity, but show very weak transfer to general reaction time. The effect is highly task-specific.
Most supplements
Claims for omega-3, ginkgo biloba, nootropic stacks, and most cognitive supplements are not backed by rigorous RCT data for reaction time specifically. Save your money for better sleep.
Trying harder / concentrating more
Effortful attention helps choice reaction time slightly but not simple reaction time. "Trying to go faster" often backfires through increased tension and anticipatory false starts.
Establish your baseline now
Test on a well-rested morning, then again after sleep deprivation, caffeine, and exercise. You will see the differences immediately.
Take the Reaction Time test