Font Size, Line Height, and the Comfort of Reading
Readers never blame your typography; they just get tired and leave, which is why type comfort is invisible and decisive.
Nobody has ever closed a browser tab thinking “the line height on that page was too tight.” That’s what makes typography such a strange problem. When a page’s argument is bad, readers can tell you why they left. When a page’s type is bad, they leave without knowing they were pushed. They felt tired, or restless, or vaguely annoyed, and attributed it to the content, or to themselves, or to nothing at all. Typographic discomfort is real, measurable in behavior, and almost never consciously perceived. You’re being judged on a channel your visitors don’t know exists.
It helps to remember that reading is a physical act before it’s a mental one. Your eyes don’t glide along a line of text; they hop, fixating on a clump of letters, jumping to the next clump, and at the end of each line executing a long leap down and back to the left margin. That return leap is the hard part. The eye has to land on the start of the correct next line, using nothing but the spacing of the lines to aim by. Most of what we call comfortable typography is just making those mechanics easy: clumps big enough to recognize instantly, leaps short enough to land reliably.
Font size governs the first half. Text that’s too small forces the eye to work at the edge of its resolving power, and the reader compensates without noticing, leaning in, squinting slightly, slowing down. On phones the failure is usually inherited rather than chosen: a size that looked generous on a designer’s monitor renders small on a dense phone screen held at arm’s length on a bus. The practical floor that’s emerged for body text on the web is around sixteen pixels, and it’s a floor, not a target. There’s also a browser behavior worth knowing: when text is smaller than that, some mobile browsers may zoom in on a form field the user taps, trying to make the input legible, and the page lurches sideways. Your typography setting off the browser’s own accessibility reflexes is a decent sign it’s too small. And small type is not a minor aesthetic sin but an exclusionary one: for older readers, whose eyes resolve less contrast and detail than they did at twenty-five, your tasteful compact text may be simply unreadable, and they are a larger share of every audience than most site owners assume.
Line height governs the second half, the return leap. If lines sit too close together, the eye’s landing zone is ambiguous; readers re-read lines or skip them, usually without registering why the paragraph felt like wading. Too far apart and the lines stop cohering into a paragraph at all, each floating alone. The comfortable range for body text sits somewhere around one and a half times the font size, looser as lines get longer, because longer leaps need bigger targets. Which points to the third variable, line length itself. A line that spans a wide monitor edge to edge can hold well over a hundred characters, and every one of those lines ends with a return leap across the full width of the screen. Sixty to seventy-five characters per line is the range readers have found comfortable for centuries, and it’s why a simple max-width on your text column is among the highest-value single lines of CSS a site can have. None of these three settings means anything alone; size, spacing, and measure form one system, and changing any one shifts what the other two should be.
Here’s the economic way to think about it. Every visitor arrives with a finite budget of reading effort, and everything on your page draws from the same account: your argument draws from it, your vocabulary draws from it, and so does the raw physical act of resolving glyphs and finding line starts. Typography determines the overhead rate. Comfortable type might spend a small share of the reader’s effort on mechanics, leaving the rest for meaning. Cramped, tiny, over-wide type can consume so much on mechanics that even willing readers exhaust themselves halfway down the page. The cruel part is the asymmetry I opened with: they’ll never bill you for it. They’ll just stop.
What convinced me of how common this is was building GazeSite, which renders every site it audits on both a desktop and a phone screen and looks at what the visitor actually sees. The pattern that keeps recurring isn’t exotic. It’s sites whose owners have plainly never read their own text on a phone. The words were written in a comfortable editor, approved in a wide design mockup, and shipped; the one format most visitors experience, a narrow column on glass under a thumb, was never once proofread. The desktop rendering and the mobile rendering of the same paragraph can be different experiences entirely, one airy and one airless, and only one of them is the product.
The honest defense of good typography is that it’s the cheapest improvement your site can get. Rewriting your copy takes days. Restructuring your pages takes weeks. Bumping body text to a legible size, opening the line height, and capping the column width takes minutes, applies to every page at once, and pays on every visit forever. Type comfort compounds like that because it isn’t content; it’s the surface every piece of content passes through.
The test costs nothing. Open your longest page on your own phone, tonight, in bed, when you’re tired, and read it all the way down. Not skim: read. If you catch yourself losing your place, pinching to zoom, or feeling a faint relief when it ends, you’ve felt what your visitors feel. They just won’t tell you. Your eyes will.
More articles
Accessibility arguments fail when framed as compliance or virtue; they succeed when framed as ordinary business investment.
Read →A page can be beautiful and still exhausting, and exhausting pages lose the sale.
Read →Scanners find markup mistakes, but only using a screen reader tells you whether a page actually works.
Read →