Blog · Research
Do dyslexia fonts improve reading comprehension — or just decoding speed?
When people switch to a dyslexia-friendly font and find reading easier, what has actually changed? They may be decoding the words faster — identifying the letters and assembling them into words without as much conscious effort. Or they may be understanding the text better — grasping the meaning, retaining the argument, connecting new information to what they already know. These two things feel similar in the moment but are measured differently, driven by different mechanisms, and affected differently by font choice. Getting the distinction right changes how you should evaluate any reading tool — including a font.
The difference between decoding and comprehension
Cognitive science draws a clear line between the two. Decoding is the low-level process of translating printed letters into words. For most adults who read fluently, decoding is automatic and unconscious — they do not notice doing it. For dyslexic readers, decoding is often slower and more effortful, even after years of literacy instruction. The letters do not map to sounds or words automatically; the brain must work harder at each step.
Comprehension sits on top of decoding. Once the word is recognised, the brain must hold it in working memory, connect it to the words that came before, build a sentence-level meaning, and then integrate that meaning into the growing mental model of the text. This process draws on vocabulary, background knowledge, inference-making, and sustained attention — none of which are directly touched by font choice.
The crucial link between the two is cognitive load. If decoding is effortful, it consumes working memory capacity that would otherwise be available for comprehension. A reader who is struggling to recognise words is a reader who has less mental bandwidth to understand what those words mean. This is why many dyslexic readers can read a page, reach the bottom, and have retained almost nothing from it — the decoding absorbed the resource that comprehension needed. (See our piece on dyslexia and working memory for a detailed look at this mechanism.)
What the research says about fonts and decoding
The strongest case for dyslexia fonts rests on decoding, not comprehension. Studies on OpenDyslexic — including work by Rello and Baeza-Yates from 2013 — found that the font reduced reading time and error rate on specific letter-confusion tasks in some dyslexic participants. The effect was not universal, but it was consistent enough to be meaningful. The weighted-bottom letter shapes appear to reduce the effort needed to distinguish b from d, p from q, and similar mirror pairs that many dyslexic readers report as genuinely confusing.
Lexend, designed by Bonnie Shaver-Troup and Thomas Jockin, takes a different approach. Rather than redesigning letter shapes, it increases letter spacing and uses a medium weight to reduce what its designers call "visual crowding" — the way closely packed letters interfere with each other's recognition. Published studies and classroom pilots found modest but real improvements in reading fluency, particularly for younger readers and those with lower baseline reading speeds. Our overview of Lexend in Chrome covers the specifics in more detail.
Atkinson Hyperlegible, designed for the Braille Institute to support low-vision readers, prioritises disambiguation — every letterform is designed so it cannot be mistaken for another. In practice, many dyslexic readers report a similar benefit to OpenDyslexic with a more conventionally attractive appearance. The research base is smaller, but the legibility logic is sound. See our guide to Atkinson Hyperlegible for a walkthrough.
In each case, the font change is acting on the decoding step: making letter recognition faster, easier, or less error-prone.
What the research says about fonts and comprehension
Here the picture is more complicated, and the honest answer is: the direct effect of font choice on comprehension is small, inconsistent, and probably indirect.
Most studies that measure comprehension alongside reading speed find one of two patterns. Either there is no significant difference in comprehension scores between the dyslexia font and the control (which typically means the font improved decoding efficiency without degrading comprehension), or there is a modest improvement in comprehension that appears to be mediated by the reduction in cognitive load rather than by any direct effect of the font on understanding.
That second pathway — font reduces decoding effort, which frees working memory, which improves comprehension — is real and worth understanding. If you were spending 40% of your working memory on recognising words in a difficult font, and a font change drops that to 20%, the freed capacity goes somewhere. Some of it goes to comprehension. A 2019 study examining font effects in readers with diagnosed reading difficulties found that comprehension scores on standardised passages improved by roughly 8-12% when readers switched from Times New Roman to a higher-legibility sans-serif, a change the authors attributed entirely to the reduction in decoding effort rather than to any direct effect on meaning-making.
The indirect mechanism: dyslexia fonts improve comprehension not by making the text easier to understand, but by making the text easier to decode — which frees the cognitive resources that comprehension actually needs. The font is not touching the comprehension process directly; it is removing a bottleneck upstream of it.
This distinction matters because it sets a realistic ceiling on what font changes can achieve. A text that is conceptually difficult — dense legal language, technical jargon, complex argument structure — will not become easier to understand because you switched to Lexend. The font will help you get through the words, but the meaning still requires the same background knowledge and inference-making ability. This is not a knock against dyslexia fonts; it is a clarification of what they actually do.
When font changes do and do not help comprehension
| Situation | Font change likely helps | Font change unlikely to help |
|---|---|---|
| Decoding is the main friction point | Yes — relieving decoding effort frees working memory for meaning | - |
| Working memory is already stretched | Yes — lower decoding cost leaves more capacity available | - |
| Text is conceptually hard for unrelated reasons | - | Comprehension difficulty is content-driven, not decoding-driven |
| Reader has strong background knowledge of the topic | Modest benefit at most | Comprehension is already efficient; bottleneck is not at decoding |
| Reader is fatigued or reading for a long stretch | Yes — reduced decoding effort delays cognitive fatigue | - |
| Text is in a second or foreign language | - | Comprehension bottleneck is vocabulary and syntax, not letter recognition |
The role of font preference and reading comfort
One factor that the controlled studies tend to underweight is reading comfort as a driver of how long someone is willing to keep reading. Comprehension in an experiment is measured over a fixed-length passage, completed in a single sitting. That is not how most adult reading happens in practice.
If a font makes reading feel less effortful, many readers will simply read more — more articles, more of a long document, more of a chapter before closing the book. This volume effect does not show up in a comprehension test, but it compounds enormously over time. A reader who finds a font comfortable enough to read 30 more minutes per day is reading perhaps 150 more hours per year. The comprehension benefit from that additional reading dwarfs any direct per-sentence font effect.
This is one reason why subjective preference data in font studies is worth taking seriously even when objective comprehension scores are flat. Studies on OpenDyslexic consistently find that readers report preferring it over default fonts even when speed and comprehension scores show no difference. The preference is not irrational — it is tracking something real about sustained reading comfort that the test doesn't capture. Similarly, the evidence on serif vs sans-serif fonts for dyslexic readers shows that preference data often predicts reading behaviour better than controlled comprehension scores do.
What actually improves comprehension for dyslexic readers
If font choice is at best an indirect lever for comprehension, what does move the needle directly?
Line spacing and letter spacing have a stronger direct effect than font choice in most studies. Appropriately wide line spacing — roughly 1.5 times the font size — reduces the visual interference between adjacent lines, which is one of the most consistent sources of comprehension disruption for readers with tracking difficulties. Letter spacing that slightly exceeds the typeface default reduces crowding. Both changes can be applied regardless of which font you use, and both show up consistently in comprehension data. Our article on line spacing and letter spacing for dyslexia covers the specific values to aim for.
Text layout also matters more than font for comprehension. Left-aligned text with a moderate line length (roughly 60-70 characters) outperforms justified text and very wide or very narrow columns on comprehension tests across multiple studies. The consistent left edge gives the eye a reliable return point and reduces the cognitive cost of tracking back to the start of a new line.
Background colour can help too, particularly for readers with visual stress — the experience of text seeming to move, flicker, or create halo effects on high-contrast white backgrounds. For those readers, switching to a warm cream, light grey, or pale blue background reduces the visual noise that competes with comprehension. See our piece on background colours for dyslexia for research-backed options.
Chunking — breaking long sentences into shorter ones, and grouping related sentences into short paragraphs — has a larger effect on comprehension than any typographic change. This is a content intervention rather than a display intervention, but it is worth knowing that the biggest comprehension gains for dyslexic readers tend to come from how text is written, not how it is rendered.
The honest case for using a dyslexia font anyway
None of the above is an argument against using a dyslexia-friendly font. The case for using one is clear: reducing decoding effort is a legitimate route to improving comprehension, even if it works indirectly. For many readers the decoding bottleneck is real and significant, and removing it produces a genuine improvement in how much they take in and retain from what they read.
The case is strongest when the reader is aware of the indirect mechanism and does not expect the font alone to close the gap. A dyslexia font is not a comprehension aid; it is a decoding aid that, by freeing cognitive resources, makes comprehension more achievable. Pairing a legible font with appropriate line spacing, a comfortable background colour, and a manageable line length will do more for comprehension than any single element on its own.
If you have not yet experimented with font changes in the browser, LexiFont applies OpenDyslexic, Lexend, Atkinson Hyperlegible, or any combination of spacing adjustments to every site you visit. The free tier is enough to run a meaningful test — pick a long article you find difficult to read, set a comfortable font and spacing, and see whether you reach the end with more retained than usual. That practical test will tell you more than the average effect in a controlled study, because it measures your specific decoding profile on your typical reading material.
If you read a lot of dense content — research papers, contracts, long-form journalism — the LexiFont Pro tier lets you set different font profiles for different sites, so you can use a more aggressive spacing on dense documents and a lighter touch on sites that are already well-designed.
Bottom line: dyslexia fonts improve comprehension indirectly, by reducing the decoding effort that competes with it. They do not improve comprehension directly. The practical implication is to pair a legible font with good spacing and background colour rather than relying on font alone — and to measure the benefit over sustained reading sessions rather than a single timed passage.
A note on what "comprehension" means in different contexts
One more nuance worth naming: comprehension is not one thing. Standardised reading comprehension tests typically measure literal recall — did the reader encode and retain specific facts from the passage? That is a narrow slice of what most people mean by understanding. Deeper comprehension involves drawing inferences, noticing contradictions, applying new information to existing knowledge, and evaluating the reliability of the source. None of these has been carefully studied in relation to font choice, and it would be surprising if font choice affected them much. They depend on the reader's prior knowledge, critical thinking habits, and engagement with the material — factors that a font simply cannot touch.
This is also why the comparison between Bionic Reading and OpenDyslexic is interesting from a comprehension standpoint. Bionic Reading, despite its marketing, appears to reduce comprehension scores on some measures, possibly because the bolding disrupts the normal visual scanning pattern that supports meaning-building. OpenDyslexic, which changes the letter shapes rather than the rendering, does not show the same comprehension decrement. Two techniques that both claim to make reading easier are doing quite different things to the comprehension process.
For a broader look at which font to start with and how to evaluate it for your own reading profile, see our guide to the best fonts for dyslexia in 2026.