By Eric, BlinkBench Founder · Last reviewed: July 2026
“What does my reaction time say about me?” is the question this whole site exists to answer honestly, and the honest answer is: less than you'd think, and less about you than about the last five minutes. Why we report the median already covers the noise inside a single run — one lapsed trial among five. Can you practise your way to a better score? covers the noise between two runs — familiarity with the interface, not a faster nervous system. This guide covers three more movers, all measured in the published literature, none of them arriving from inside the test at all: what you drank, how you slept, and what time it is.
Caffeine moves it, by a modest and well-measured amount
A 2025 meta-analysis pooled 31 placebo-controlled trials — 1,455 healthy adults — testing caffeine against a dummy pill on behavioral attention tasks. Caffeine won on reaction time, with an effect size of g = 0.28, and the benefit scaled with dose: 200 mg or more moved the number further than a smaller dose, in a roughly straight line.
That g is a standardized effect size — a mean difference expressed in pooled standard deviations, not milliseconds — and this page won't convert it into one. Doing that honestly needs the actual spread of the trials behind it, which the published figure doesn't supply, and inventing a number to fill that gap is exactly the move the reaction time test's own apparatus statement refuses to make for hardware. A g of 0.28 is real and it is modest — smaller than either of the next two factors.
A bad night moves it more, in the other direction
A 2010 meta-analysis pooled 26 studies — roughly 1,533 participants — measuring cognitive performance after 24 to 48 hours of short-term sleep deprivation. Of every domain it tested, simple reaction time and vigilance were hit hardest: an effect size of −0.732 on reaction time, and −0.762 on outright lapses (misses), both noticeably larger than the caffeine effect above, and in the direction you'd expect.
Put the two side by side and the standardized units line up cleanly even though the raw numbers don't: short-term sleep loss degrades the same kind of measurement by roughly twice what a dose of caffeine buys back. Coffee does not undo a bad night; it makes a smaller correction in the opposite direction.
The clock moves it too, and the literature can't agree which way
Here the honest answer changes shape. A 2024 systematic review collected every study it could find on time-of-day effects on simple reaction time — eight of them. Four found people fastest in the evening, by roughly 9 to 13% over other times tested. Two — both in athletes — found people fastest in the morning, with performance declining by around 34% as the day wore on. Two found no time-of-day effect at all.
That is not a small range narrowing in on a true answer; it is eight studies pointing in three different directions. Caffeine and sleep deprivation both get a number and a sign above, because the published effect sizes agree on direction even where they differ in exact magnitude. Time of day doesn't clear that bar, so it gets named as a factor and nothing more — no number, no “test in the morning for your best score.” A guide that manufactured confidence the studies themselves don't have would be doing exactly what this site criticizes uncited percentiles for.
None of this is a claim about who you are
Five things now sit on this list of what moves a reaction-time number besides the person: a single lapsed trial, familiarity from a prior run, caffeine, sleep debt, and — named but unquantified — the hour on the clock. Every one of them is a state, not a trait. None of them is raised by practice in any sense this site is willing to sell you, and this page isn't going to tell you to caffeinate before testing or schedule your attempt for a particular hour — that would turn a measurement into a lifestyle tip, and the tip would be resting on effect sizes this modest.
What the number is actually good for is the same thing it was always good for: a reading, taken once, under stated conditions, of an instrument that includes your display, your input device, and — this page's addition — the particular five minutes you happened to be in when you took it.
Sources
- Kløve, K., & Petersen, A. (2025). A systematic review and meta-analysis of the acute effect of caffeine on attention. Psychopharmacology (Berl), 242(9), 1909–1930.
- Lim, J., & Dinges, D. F. (2010). A meta-analysis of the impact of short-term sleep deprivation on cognitive variables. Psychological Bulletin, 136(3), 375–389.
- Munnilari, M., Bommasamudram, T., Easow, J., Tod, D., Varamenti, E., Edwards, B. J., Ravindrakumar, A., Gallagher, C., & Pullinger, S. A. (2024). Diurnal variation in variables related to cognitive performance: a systematic review. Sleep and Breathing, 28(1), 495–510.