I started testing my reaction time every morning after reading that it was one of the best single-number proxies for cognitive health. Not fitness. Not IQ. Reaction time. The research made a compelling case. So I ran the experiment for 30 days — same time each morning, same protocol, tracking everything I did the night before and that morning. What I found was messier and more interesting than I expected.
01 /Why reaction time matters more than most people think
Here's the thing about reaction time that surprised me when I first dug into the literature: it's not just a measure of how fast your fingers move. It's a measure of how efficiently your entire nervous system is processing information.
The signal has to travel from your eyes to your visual cortex, get interpreted, trigger a decision, and then send a motor command down to your hand. All of that happens in under 300 milliseconds for most healthy adults. When something slows that chain down — poor sleep, alcohol, stress, aging — every step gets a little worse.
What I didn't fully appreciate before running this experiment was how sensitive the measure is. We're talking about changes of 20-30 milliseconds mattering. That's the difference between an average 25-year-old and an average 35-year-old in population norms. Small numbers, but they're not random noise.
02 /What actually made me slower
Alcohol was the most dramatic. Even two drinks the night before produced a measurable slowdown the next morning — typically 30-50ms compared to my baseline. I'd read that alcohol impairs reaction time during intoxication, but I hadn't expected the hangover effect to be this consistent.
Poor sleep was second. Anything under 6.5 hours and I could see it in the numbers. Not always — some nights of bad sleep didn't seem to affect the next morning. But the correlation was strong enough that I stopped being surprised.
Stress was harder to quantify, but on days following high-stress periods — a difficult conversation, a deadline — my scores were consistently worse. Whether that's the stress itself, the poor sleep that often accompanies it, or some combination, I couldn't fully separate.
Late-night screen time correlated weakly with worse morning scores, but I couldn't tell if that was the screens or just the fact that late screens meant less sleep.
03 /What actually made me faster
Aerobic exercise was the most consistent positive effect I found. Not on the same day — on the day after a hard cardio session, my scores were noticeably better. Not dramatically, maybe 15-20ms, but it was consistent enough across the 30 days that I started to trust it.
This matches what the research says. A 2020 meta-analysis found that regular aerobic exercise improves simple reaction time by an average of 22ms. That's meaningful — it's the equivalent of being cognitively 5-10 years younger on this particular measure.
Caffeine helped, but less than I expected. My morning coffee seemed to restore my baseline after a rough night, but I didn't find consistent evidence that it improved my performance above my well-rested baseline. Which is exactly what the literature would predict — caffeine mostly reverses caffeine withdrawal, not a genuine enhancement on top of a normal baseline.
The biggest surprise: consistent sleep schedule mattered more than total sleep hours. The days after I'd slept 7 consistent hours for several days in a row were better than the days after a catch-up 9-hour sleep following a bad week. Your circadian rhythm seems to matter, not just the quantity.
04 /So what does this mean for your score?
If you've taken a reaction time test and you're not happy with your result, the first question worth asking is: what were the last 24 hours like? One bad night's sleep is enough to add 30-50ms to your reaction time. That's enough to move you from 'above average' to 'below average' on most scoring curves.
The more interesting question is what your baseline looks like — not after coffee, not after a bad night, but after three good consecutive nights of sleep, moderate exercise, and no alcohol for 48 hours. That's when you're seeing something closer to your actual underlying cognitive speed.
And that number — the real baseline — tends to be more stable than people expect. It changes slowly with age, and changes meaningfully with sustained lifestyle habits. It doesn't jump around based on daily mood. Which makes it a genuinely useful signal if you test it under consistent conditions over time.
05 /One thing I didn't expect
Thirty days in, my average reaction time had improved by about 25ms compared to day one. Some of that is almost certainly practice — learning the test, learning to anticipate. But some of it might be real. I'd been sleeping more consistently, exercising more regularly, and drinking less, partly because I was tracking the outputs and could see the effects.
That might be the most honest thing I can say about reaction time testing: the value isn't necessarily the number. It's that having a number makes you take the inputs seriously in a way that abstract advice about sleep hygiene never quite does.
When you can see that two drinks cost you 40ms the next morning, the cost of those drinks becomes concrete. That's worth something.