Readable measure

Claim: body text has a line length of about 45 to 75 characters, a size in the fluent range, and line spacing of at least 1.5.

Grade: measured for a size in the fluent range. A line of 45 to 75 characters: supported with caveat, best for understanding and preference but not for speed. Line spacing of at least 1.5: a convention.
Evidence: typography and layout (strands 1 to 3), baseline.
Check: readable-measure (exact, on the site's stylesheet): a measure of 45 to 75 characters, body text of 16 px or more, line height of 1.5 or more. figure-width (a warning): a figure drawn in characters stays within 60 columns, so it fits where its text fits.
Reader styles: none override it; letter spacing may be raised for readers with dyslexia.

What it asks

45 to 75 characters per line for a single column, measured on the rendered page.
Body size large enough to sit well inside the fluent range on the reader's device; 18 px is a safe default on screen.
Line height of 1.5 or more, and text that survives the WCAG 1.4.12 spacing overrides.
Left-aligned text, no long runs of italics.

Why

Reading speed holds across a tenfold range of print sizes and falls below it (Legge 2011). Comprehension was higher at 18 pt than at smaller sizes in a sample of 104 (Rello 2016). Common serif and sans faces read equally fast on screen (Bernard 2002).

Limits

Speed and preference disagree on line length: about 100 characters is read fastest on screen, about 55 is understood best and preferred (Dyson 2004). Fonts designed for dyslexia do not help (Wery 2017; Kuster 2018); extra letter spacing does help dyslexic children read accurately (Zorzi 2012).

The upper bound is contested in practice: WCAG's AAA criterion allows 80 characters, 40 in CJK (W3C 2023a), and the US Web Design System 90, with 66 as its target (USWDS 2026). Neither page cites a study: both are convention, and the data above split speed from understanding (Dyson 2004). Line spacing of 1.5 comes from practice guides as well (USWDS 2026, convention); mid-range spacing made no difference to comprehension in one sample (Rello 2016).