IQ and Reading: Does Reading Actually Raise Intelligence?

Reading raises some forms of intelligence and barely touches others. It reliably builds verbal IQ, vocabulary and background knowledge; its effect on raw reasoning power, the part psychologists call g, is real but modest. That distinction is the whole story, and most advice on the topic skips it.

The numbers behind the strongest claim are striking. In a sample of 1,890 identical-twin pairs tracked from ages 7 to 16, differences in reading ability predicted later differences in intelligence — and the effect was not confined to verbal tests (Ritchie, Bates & Plomin, 2015). Because identical twins share their DNA, that design strips out genetics and points to reading itself as a candidate cause. As an interpretation rather than a finding of my own, I read that study as the best single piece of evidence that reading does something to the mind beyond signalling who was already bright. It is also, as you will see, contested.

Where reading helps most maps neatly onto what the CMIAS framework calls Cross-Domain Transfer — the capacity to carry a pattern learned in one field into an unrelated one. Wide reading is, in effect, transfer practice at scale.

Does reading raise IQ?

Short answer: yes, but unevenly, and the size of the effect depends entirely on which slice of intelligence you measure. The single most informative study used identical twins precisely to settle the question of cause. Ritchie, Bates and Plomin (2015) tested 1,890 monozygotic twin pairs at five ages and asked a sharp question — when one twin pulls ahead of the genetically identical co-twin in reading, does that twin later pull ahead in intelligence? The answer was yes, and the link was not restricted to verbal tasks. Because the twins share their genes, the difference had to come from the non-shared environment: different teachers, different books, different reading habits.

That is unusually clean evidence in a field full of confounds. It does not prove that an hour of reading buys a fixed number of IQ points — no credible study claims that — but it does suggest reading ability feeds back into general cognition during the school years.

Here is the honest qualification most coverage leaves out. A 2017 reanalysis of the very same data by Bailey and colleagues argued that some of the apparent reading-to-intelligence effect could be an artifact of stable underlying factors influencing both reading and IQ throughout development, rather than reading acting as a clean cause. Ritchie's team did real work to rule out obvious confounds; the reanalysis is a reminder that "cause" in developmental data is rarely airtight. Both can be true at once: reading probably contributes, and we are not certain how much.

Which kind of intelligence reading builds

The key split is between fluid and crystallized intelligence. Fluid intelligence is on-the-spot reasoning with novel material; crystallized intelligence is the accumulated store of knowledge, vocabulary and verbal skill. Reading pours almost directly into the crystallized side. The difference between the two matters so much for this topic that it is worth understanding the distinction between fluid and crystallized intelligence before drawing conclusions about what any reading habit can do for you.

A reader's vocabulary, general knowledge and verbal reasoning all climb with print exposure. These are exactly the abilities that load onto verbal IQ subtests, which is why heavy readers tend to score well on the verbal half of an assessment. The relationship between reading and the verbal versus nonverbal split in IQ testing is strong and well-replicated.

Fluid reasoning is a harder sell. Reading exposes you to arguments, structures and analogies, which plausibly sharpens reasoning — and the twin study's non-verbal effect hints at it — but the gains are smaller and noisier than the vocabulary gains. If someone promises that a reading habit will lift your matrix-reasoning score the way it lifts your vocabulary, treat that with skepticism. The cleanest, largest, most reliable effect of reading is on what you know and the words you have for it.

To see where your own verbal and reasoning ability sits relative to population norms, the Advanced IQ Test measures open-answer reasoning alongside vocabulary and logic in a single 40-minute session.

Why reading is the vocabulary engine

Here is a number that reframes the whole debate. Cunningham and Stanovich (1998) found that children's books contain roughly 50 percent more rare words per thousand than adult prime-time television, and substantially more than the conversation of even college-educated adults. Speech, it turns out, is lexically thin. The sophisticated words that mark a large vocabulary mostly do not appear in talk; they live in print.

That single fact explains why reading volume tracks vocabulary so tightly. If you want exposure to "ostensible," "recalcitrant" or "asymptotic," you will encounter them on a page long before you hear them at dinner. Spoken language, even fairly educated spoken language, recycles a small core of common words. Print does not.

The mechanism connects to working memory too, because comprehending complex sentences while holding earlier clauses in mind is a memory-intensive act. The link between working memory and measured IQ is one reason skilled readers and high scorers overlap: parsing a dense paragraph and solving a reasoning item draw on some of the same machinery.

"The reason 'just read more' is decent advice is almost embarrassingly mundane: books are where the rare words are. You cannot acquire a vocabulary you never meet, and most people only meet sophisticated language on a page."

— Adam Imran, MS Clinical Psychology · DesperateMinds

The Matthew effect: the rich get richer

Keith Stanovich gave this loop a name in 1986, borrowing from the Gospel of Matthew: to those who have, more is given. Applied to reading, the dynamic is brutal in its simplicity. A child who decodes text slightly faster reads slightly more, meets more words, builds more knowledge, and so finds the next book slightly easier — which makes them read still more. A child who struggles early reads less, meets fewer words, and falls steadily behind. Small initial gaps widen into chasms.

Mol and Bus (2011) put numbers to the upward spiral. Their meta-analysis pooled 99 studies covering 7,669 readers from preschool through university and found moderate-to-strong correlations between print exposure and both reading comprehension and technical reading skill — correlations that grew stronger with age. The longer the loop runs, the more it compounds. That is the encouraging half and the discouraging half of the same finding: reading habits are self-reinforcing in both directions.

What does this mean for an adult who reads little and wants to change that? The compounding logic still applies. The first month is the hardest because vocabulary and stamina are lowest; every month after gets easier as the base grows. For a fuller picture of which habits genuinely move cognitive performance, our guide to evidence-based ways to increase IQ places reading in context against sleep, education and the rest.

The causal debate, honestly stated

This is where most internet coverage gets it wrong. Articles tend to pick one of two extremes — "reading makes you smarter, period" or "it's all just correlation, smart people happen to read" — and neither survives contact with the actual literature.

The correlation is genuine and large: more reading goes with higher measured intelligence almost everywhere you look. The question is direction. Three explanations compete. Reading might cause intelligence gains. Intelligence might cause more reading. Or a hidden third factor — say, a curious temperament, or a home full of books and educated parents — might drive both. The honest answer is that all three operate, and disentangling their shares is genuinely hard.

The twin design was built to cut through this. By comparing genetically identical co-twins, Ritchie, Bates and Plomin (2015) removed the genetic third factor and still found reading predicting later intelligence. That is strong. But Bailey and colleagues' 2017 reanalysis showed that even within twin pairs, stable shared influences can masquerade as a reading effect. My own reading of the back-and-forth: the causal contribution of reading is probably real and probably smaller than the headline twin result implies. Anyone who quotes the 2015 study without mentioning the 2017 critique is giving you half the picture.

One reasonable caution about the cited work itself. Ritchie's team measured "reading ability" rather than "reading volume," and explicitly noted the effect was not explained by reading exposure differences — meaning the study is more about the skill of reading than the simple act of reading widely. That nuance often gets flattened in popular summaries into "reading books raises IQ," which is not quite what the data said.

Can adults still gain from reading?

Yes — and the gains skew toward crystallized intelligence, which is the good news, because crystallized ability holds up and even rises across most of adulthood while fluid reasoning slowly declines. Stanovich, West and Harrison (1995) showed that print exposure predicts knowledge growth and maintenance across the lifespan, independent of raw ability. Older adults who read widely retain and extend their knowledge base in ways non-readers do not.

There is a tangible career angle here that often goes unmentioned. Vocabulary and verbal facility correlate with occupational status and earnings, and the relationship between measured intelligence and income runs partly through exactly the verbal and knowledge channels that reading feeds. The adult who keeps reading is, in a slow and unglamorous way, maintaining an asset.

The reading slump and what it costs

The trend is moving the wrong way. The National Endowment for the Arts found that 48.5 percent of US adults read at least one book for pleasure in 2022, down from 52.7 percent in 2017 and 54.6 percent a decade earlier. Federal time-use data put daily reading "for personal interest" at roughly 15 to 16 minutes per person across 2017–2023. A 2025 analysis of two decades of the American Time Use Survey reported a steady decline of about 3 percent per year in daily reading for pleasure (Bone et al., 2025).

Buried in that 2025 data is a counterintuitive twist worth flagging directly: among people who still read, average time spent reading actually rose slightly, from about 1 hour 23 minutes a day in 2003 to 1 hour 37 minutes in 2023. The population is polarizing. Fewer people read at all, but the committed read more. The Matthew effect, scaled up to a whole society.

If reading even modestly supports verbal intelligence and knowledge maintenance, a society reading less is quietly drawing down a cognitive resource — and concentrating it among a shrinking group. Whether that shows up in national test scores over the coming decade is an open empirical question, and one I would not bet against.

Conclusion

Reading is not a magic IQ pill, and the people selling it as one are overpromising. What it reliably does is build the verbal half of intelligence — vocabulary, knowledge, comprehension — through the simple fact that books hold words and ideas that conversation never surfaces. Its effect on raw reasoning is smaller and still argued over. But "build a large vocabulary, a deep store of knowledge, and the verbal reasoning that rides on both" is hardly a consolation prize. If you want one habit with a strong evidence base and almost no downside, this is the one — just hold the expectation steady and let the compounding do the rest.

Frequently Asked Questions

References