Does Dyslexia Affect IQ Tests?

Discover how dyslexia impacts working memory and processing speed to suppress standard IQ scores. Read our full guide and try the RIOT IQ test today!

Dr. Russell T. WarneChief Scientist

Dyslexia and IQ have a complicated relationship in both the research literature and in public understanding, and the two tend to get conflated in ways that do real harm to how individuals with dyslexia are assessed, classified, and supported. The short answer is yes, dyslexia does affect IQ test performance, but not in the direction or for the reasons most people assume. The effect is not on general cognitive capacity; it's on the specific subtests that happen to tap the same phonological and processing speed systems that dyslexia impairs. Understanding that distinction is one of the more practically important things I can explain in the context of cognitive assessment.

What Dyslexia Actually Is

Dyslexia is a specific reading difficulty rooted in a phonological processing deficit, a difficulty in mentally manipulating the sound structure of language that makes decoding written words effortful and inaccurate. Dyslexia is specifically a phonological processing difficulty, not a general cognitive impairment. The phonological deficit is detectable before formal reading instruction begins and persists into adulthood, even in individuals who have learned to read fluently through compensatory strategies.

The neurological basis is well-established. Functional MRI studies consistently show that dyslexic readers differ from typical readers in three main areas: the left occipitotemporal cortex (the visual word form area) is under-activated; the left temporoparietal cortex shows reduced activation during phonological tasks; and compensatory over-activation appears in right-hemisphere and frontal regions. Critically, dyslexic readers show similar neural patterns regardless of IQ level; high-IQ and average-IQ dyslexic readers are neurologically indistinguishable when reading. The phonological deficit is independent of general intelligence.

This neurological independence is the foundation for everything else I want to say in this article.

Why the Old Diagnostic Model Was Problematic

For most of the 20th century, the standard diagnostic approach to dyslexia required a significant discrepancy between IQ and reading scores — reading performance substantially below what IQ would predict. This "discrepancy model" created a circular problem that damaged the identification and support of countless children: children with dyslexia and lower-than-average IQ scores were often not classified as learning disabled and therefore not eligible for special educational services, because their reading difficulties looked proportionate to their IQ rather than discrepant from it. In other words, the model implicitly assumed that dyslexia was only real if the person was otherwise "smart enough" — a premise that neuroscience has since dismantled entirely.

In October 2025, the International Dyslexia Association updated its definition and diagnostic framework to move away from this model. The new definition focuses on ongoing reading and spelling difficulties that are unexpected and not explained by intelligence or poor instruction, allowing identification much earlier — beginning around age five using tools like the CTOPP-2 and AI-based eye-tracking — rather than waiting for a measurable reading failure to emerge. This change reflects a scientific consensus that has been building for over two decades.

Where Dyslexia Actually Shows Up on an IQ Test

The critical psychometric finding is this: dyslexia doesn't lower general intelligence, but it does produce a characteristically uneven cognitive profile on a standard battery that can depress the Full Scale IQ score if the examiner doesn't interpret it correctly.

A 2025 WAIS-IV study of 36 Italian university students with dyslexia compared to 35 matched controls found that while dyslexic adults showed lower Full Scale IQ scores than controls, their scores remained within the average range. The critical finding was in the index profile: dyslexic adults showed significant deficits in the Working Memory Index and Processing Speed Index, but performed similarly to controls on the Verbal Comprehension Index and Perceptual Reasoning Index. Logistic regression identified WMI and PSI as the most reliable predictors of dyslexia, showing good predictive value in discriminating between adults with and without the condition.

Research on children with developmental dyslexia using the WISC-III found significant deficits in subtests relying on verbal abilities, processing speed, and working memory, with larger effect sizes in Information, Arithmetic, and Digit Span. This pattern of stronger verbal comprehension and perceptual reasoning, weaker working memory and processing speed, is now recognized as the characteristic dyslexia profile on Wechsler batteries.

Why specifically these domains? Processing speed subtests like Coding and Symbol Search require rapid scanning and transcription of symbols, the same rapid phonological and visual processing that dyslexia impairs. Working memory subtests like Digit Span require holding and manipulating verbal sequences in mind, which draws on the same phonological loop system that dyslexia taxes directly. By contrast, matrix reasoning, block design, and verbal analogies draw much less on phonological processing, which is why those subtests tend to be preserved.

The Score Suppression Effect

The consequence of this uneven profile is what researchers and clinicians call score suppression: the FSIQ is pulled downward by the WMI and PSI deficits even when the person's actual reasoning ability — as measured by the Verbal Comprehension and Perceptual Reasoning indices — is fully intact.

Research has estimated that IQ tests can underestimate true intellectual capacity by 5 to 15 points or more in individuals with learning disabilities, depending on the disability and the test format. A landmark study by Siegel found that children with dyslexia scored an average of 12 points lower on processing speed and verbal subtests compared to accommodated conditions. That is a clinically significant difference; a 12-point suppression can be the difference between a score that qualifies for support and one that doesn't.

Timed subtests compound the disadvantage particularly severely. The disability, not cognitive ability, determines how many items are completed — meaning the score is reflecting processing speed impairment rather than reasoning capacity. A student with dyslexia who scores 95 on a test relying heavily on timed reading may actually have a reasoning profile that would score 107 or higher under appropriately accommodated conditions.

What a Proper Assessment Should Do Differently

The solution isn't to avoid assessing people with dyslexia — it's to interpret their profile correctly rather than collapsing it into a single number that misrepresents both their abilities and their difficulties.

For individuals with dyslexia, the General Ability Index (GAI) — a composite that includes only Verbal Comprehension and Perceptual Reasoning, omitting Working Memory and Processing Speed — provides a more accurate estimate of general reasoning ability than the Full Scale IQ. Large-sample profiling of thousands of children with verified dyslexia found that the great majority sit at average intellectual functioning with preserved reasoning, and concluded that profile-based interpretation should supplement global indices rather than replace them.

The appropriate assessment approach is to examine the index profile for the characteristic pattern — preserved VCI and PRI alongside depressed WMI and PSI — rather than reading the FSIQ as the primary interpretive anchor. Large discrepancies between indices suggest specific learning disorders rather than global intellectual disability, and a difference of 20 or more points between the highest and lowest index scores is a clinically significant flag that the FSIQ is not an accurate summary of the person's cognitive capacity.

Tests and subtests with minimal reading or phonological decoding demands — matrix reasoning, block design, spatial relations — provide a better estimate of fluid intelligence for individuals with dyslexia than verbally loaded or heavily timed formats.

The Dyslexia–Intelligence Independence: What the Data Shows

I want to be clear about what the research establishes, because this is where popular understanding tends to be most distorted. Dyslexia is not associated with lower general intelligence. It co-occurs across the full ability range — from intellectual disability through profound giftedness — and reading difficulty is not a proxy for lower intelligence. The central finding from both neuroimaging and large-sample profiling is consistent: the phonological deficit that defines dyslexia operates independently of the reasoning and knowledge systems that IQ tests are primarily designed to measure.

What dyslexia does affect is the efficiency of certain cognitive operations, particularly rapid phonological processing and working memory for verbal sequences, in a way that happens to overlap with specific subtests on standard batteries. That overlap creates a misleading suppression effect that a single FSIQ number cannot distinguish from genuine general cognitive difficulty. Recognizing that distinction is what separates an accurate cognitive assessment from a misleading one.

The Takeaway

Dyslexia does affect IQ test performance — specifically by depressing working memory and processing speed indices while leaving verbal comprehension and perceptual reasoning largely intact. The result is a characteristically uneven profile that, if read correctly, is actually informative about the nature of the condition. If read incorrectly — collapsed into a single FSIQ and treated as a summary of general ability — it produces a score that systematically underestimates reasoning capacity and misleads the people relying on it.

The practical implications are direct: any cognitive assessment of a person with dyslexia should report index scores separately, flag significant inter-index discrepancies, interpret the GAI as the primary measure of reasoning ability where WMI and PSI are substantially depressed, and avoid making high-stakes decisions based on an FSIQ that includes substantially impaired subtests alongside preserved ones.

If you want to understand your own cognitive profile across the domains that dyslexia does and does not affect, the RIOT gives you a domain-level report that shows each index separately — so that what you're strong in and what you find difficult are both visible, rather than averaged into a single number that obscures the difference.

References

Cogn-IQ.org. (2026). Dyslexia and IQ: Not Low Intelligence. https://www.cogn-iq.org/blog/dyslexia-and-iq/
What's Your IQ. (2026). Dyslexia and IQ: High Intelligence with Reading Challenges. https://whats-your-iq.com/en/articles/learning-disabilities/dyslexia-iq-cognitive-strengths
What's Your IQ. (2026). Understanding Learning Disabilities: How They Affect IQ Testing. https://whats-your-iq.com/en/articles/learning-disabilities/learning-disabilities-impact-iq-testing
NIH / NICHD. (2011). NIH-Funded Study Finds Dyslexia Not Tied to IQ. https://www.nichd.nih.gov/newsroom/releases/110311-dyslexia-IQ
PubMed / MDPI Brain Sciences. (2025). WAIS-IV Cognitive Profiles in Italian University Students with Dyslexia. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12387714/
PubMed. (2013). WISC-III cognitive profiles in children with developmental dyslexia: specific cognitive disability and diagnostic utility. https://pubmed.ncbi.nlm.nih.gov/24222530/
Verdant Psychology. (2025). How is the Latest Research on Dyslexia Changing Diagnosis Today? https://www.verdantpsychology.com/blog-resources/latest-research-on-dyslexia
PubMed Central. (2023). Cognitive Profile Discrepancies among Typical University Students and Those with Dyslexia and Mixed-Type Learning Disorder. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10671892/