Every blog on the internet says a small "5 min read" at the top. Nine times out of ten, that number was calculated by dividing the word count by 250. It's a good default and it's also almost always wrong for you, personally.
Real reading speeds vary from about 150 words per minute (technical material, non-native language) to over 600 wpm (skimming a familiar topic). A single 250 doesn't fit any of those. Here's what actually decides the number — and how to give people a more honest estimate.
What the research actually says
A landmark 2019 meta-analysis in the Journal of Memory and Language pooled 190 studies and found the average adult silent reading speed for non-fiction English is 238 wpm, and for fiction 260 wpm. But the standard deviation was huge: roughly 60 wpm. That means "average" covers people reading at anywhere from 180 to 300 wpm — a 65% range.
Younger adults (18–30) read about 15% faster than older adults (60+). Native English speakers read about 30% faster than non-natives at the same proficiency level. Screens are consistently 10–30% slower than print for the same content — a robust finding across four decades of studies.
Content type matters more than anything
The single biggest variable isn't age or medium — it's what you're reading. Casual fiction: 250–300 wpm. News: 200–250. Business/marketing copy: 220. Legal or medical text: 100–150, and re-reading is common. Code, formulas, or dense reference material: 80–120 wpm, sometimes less. Poetry: sometimes 40, because people re-read every line.
If a page is 30% code samples and 70% prose, the naive count is way off. A better estimate weights those chunks: prose at 240 wpm, code at 100 wpm, images at a flat 12 seconds each. That's how a good reading-time estimator gets close to reality.
Why writers should show reading time at all
Nielsen Norman Group ran an eye-tracking study in 2011 that showed users spent 22% longer on articles when a reading estimate was shown. Not because they read faster — because they made a more confident choice to start reading. The label reduces uncertainty about the commitment.
The other side of it: an honest estimate builds trust. If your "3 min read" is actually a 12-minute technical deep dive, readers bounce and remember you as clickbait. Slight overestimates (say the article is 8 minutes when it's 6) actually convert better — people feel like they finished faster than expected.
Reading speed changes throughout the day
Chronotype research from University of Bologna (2020) found reading speed peaks 2–4 hours after waking for most people, drops by roughly 15% in the mid-afternoon (the well-known slump), and recovers slightly in the early evening. If you're an evening person, everything shifts by about three hours.
Fatigue matters more than caffeine. Two hours of continuous screen reading drops speed by 10–20% and comprehension by more. If you're publishing long articles, sub-headers and section breaks let people take micro-breaks without losing their place — which measurably increases finish rates.
How to estimate honestly
Take the word count, divide by 240 (a more honest average than 250), and round up. For content with code, add 30 seconds per code block over 10 lines. For image-heavy content, add 10 seconds per image (people pause to look). Round to the nearest minute. If it's under a minute, just say "1 min read" — nobody trusts "37 seconds".
Bluebird's Reading Time Calculator does all of this automatically, in your browser, so you can paste an article draft and get a realistic estimate without shipping the text anywhere. Useful before you hit publish, and useful for setting your own daily reading budget.
