I tested the same woman twice, twenty-three years apart.

The first time, she was a twenty-two-year-old college senior referred for a learning disability evaluation. Her processing speed was exceptional, in the 95th percentile. Her working memory was strong. Her vocabulary, solid but unremarkable, fell at the 60th percentile. Her Full Scale IQ came in at 118.

The second time, she was forty-five, referred by her employer after she expressed concern about “losing her edge.” She was convinced something was wrong. Tasks that had once felt effortless now required deliberate concentration. She could not hold as many things in mind simultaneously. She felt slower.

Her test results told a more interesting story than she expected. Her processing speed had indeed declined, dropping to the 70th percentile. Her working memory was slightly lower. But her vocabulary had climbed to the 90th percentile. Her fund of general information was substantially richer. Her verbal reasoning, the ability to draw inferences from language and apply accumulated knowledge, was stronger than it had ever been. Her Full Scale IQ was 121.

She was not losing her mind. She was aging. And aging, when it comes to intelligence, is not the simple story of decline that most people imagine. It is a story of trade-offs, compensations, and surprising gains that unfold across decades in patterns that neuroscience is only now beginning to fully understand.

You Do Not Have One Intelligence. You Have Many.

The most important concept for understanding how IQ changes with age was articulated by psychologist Raymond Cattell in the 1940s and refined by his student John Horn in the 1960s. They proposed that what we call “intelligence” is not a single entity but at least two broad, partially independent systems.

Fluid intelligence encompasses the ability to reason abstractly, solve novel problems, identify patterns, and think logically in situations where prior knowledge offers little help. It is the cognitive capacity you bring to entirely new challenges. On IQ tests, it is measured by tasks like matrix reasoning, where you must identify the rule governing a sequence of abstract shapes, or by figure weights, where you must determine which combination of shapes balances a scale.

Crystallized intelligence encompasses the accumulated knowledge, vocabulary, facts, skills, and learned procedures you have acquired over a lifetime. It is everything you know and can retrieve. On IQ tests, it is measured by vocabulary definitions, general information questions, verbal comprehension, and the ability to explain similarities between concepts.

These two systems follow dramatically different trajectories across the lifespan, and understanding that divergence is essential to interpreting what any IQ score means for a person of any age.

The Surprising Timeline of Cognitive Peaks

For decades, the conventional wisdom in psychology was simple: cognitive abilities peak in late adolescence or early adulthood, plateau briefly, and then decline steadily. This narrative turned out to be substantially wrong.

In 2015, Joshua Hartshorne at MIT and Laura Germine at Massachusetts General Hospital published a landmark study in Psychological Science that transformed our understanding of cognitive aging. Using data from 48,537 online participants combined with comprehensive normative data from standardized IQ and memory tests, they mapped the age of peak performance across dozens of cognitive tasks.

What they found was not a single peak followed by uniform decline. It was a staggered cascade of peaks scattered across the entire adult lifespan, each cognitive ability following its own developmental timeline.

Processing speed, the raw quickness with which you can scan visual information and execute simple cognitive operations, peaks in the late teens and early twenties and then declines relatively rapidly. This is the ability that makes young video game players so formidable and young air traffic controllers so quick on their feet.

Working memory, the ability to hold and manipulate information in mind simultaneously, climbs into the late twenties to early thirties before beginning a slow decline. This is the cognitive workspace where you juggle multiple considerations, perform mental arithmetic, and follow complex instructions.

The ability to recognize faces, surprisingly, does not peak until the early thirties, a finding that Germine had first published in 2011 and that did not fit neatly into existing theories.

Social cognition, the ability to detect other people’s emotional states from subtle cues, peaks even later, in the forties to fifties, and does not begin to significantly decline until after sixty. This finding, as Germine noted, fits “absolutely no theoretical framework that we currently have.”

And vocabulary, the quintessential measure of crystallized intelligence, showed the most striking pattern of all. While data from the original Wechsler IQ test standardization suggested vocabulary peaked in the late forties, Hartshorne and Germine’s new data pushed that peak into the late sixties or early seventies. When they investigated this discrepancy using decades of General Social Survey data, they discovered the peak had been shifting later with each generation, likely reflecting increased education, more cognitively demanding careers, and greater lifelong access to reading material.

As Hartshorne summarized the findings: “At any given age, you are getting better at some things, you are getting worse at some other things, and you are at a plateau at some other things. There is probably not one age at which you are peak on most things, much less all of them.”

What This Means for IQ Scores at Different Ages

The Cattell-Horn fluid-crystallized distinction has profound implications for how IQ scores should be interpreted at different life stages, and it is a distinction I find myself explaining to patients, attorneys, and families constantly.

When I test a child, I am primarily measuring developing cognitive architecture: how quickly their brain processes information, how much they can hold in working memory, how effectively they can reason with novel material. Crystallized intelligence at age seven reflects only seven years of accumulated learning. At that age, fluid abilities dominate the overall score, and the score is more sensitive to developmental rate than to accumulated knowledge.

When I test a young adult in their twenties, I am capturing something close to peak fluid intelligence. Their processing speed is at or near its lifetime maximum. Their working memory is approaching its ceiling. Their abstract reasoning is at full power. But their crystallized intelligence is still building. They simply have not lived long enough to accumulate the depth of knowledge, vocabulary, and real-world pattern recognition that will come with experience.

When I test someone in their fifties or sixties, I am often measuring a brain that has traded speed for depth. Their processing speed and working memory may have declined by a meaningful margin compared to their twenties. But their vocabulary, general knowledge, verbal reasoning, and accumulated wisdom have had decades to develop. On IQ tests that weight crystallized abilities heavily, like the Verbal Comprehension Index of the WAIS-IV, middle-aged and older adults often score as well as or better than they would have at twenty-five.

This is why a single IQ number, divorced from the subscale pattern and the person’s age, can be deeply misleading. A Full Scale IQ of 110 in a twenty-year-old and a Full Scale IQ of 110 in a sixty-year-old can represent very different cognitive profiles with very different practical implications.

The Neuroscience of Cognitive Aging

The divergent trajectories of fluid and crystallized intelligence are not just psychometric abstractions. They reflect real, measurable changes in brain structure and function.

Fluid intelligence depends heavily on the prefrontal cortex and a distributed frontoparietal network that supports working memory, abstract reasoning, and cognitive control. These regions are among the most metabolically active in the brain and, critically, among the most vulnerable to age-related change. Research using the Cambridge Centre for Ageing and Neuroscience dataset has demonstrated that age-related declines in fluid intelligence are partially mediated by reduced responsiveness of frontoparietal regions during novel problem-solving. The brain’s “executive control center” literally becomes less reactive with age.

White matter integrity, the health of the myelinated fiber tracts that connect distant brain regions and enable rapid information transfer, deteriorates progressively from midlife onward. Processing speed, which depends on the efficiency of signal transmission across long-distance neural pathways, is particularly sensitive to this white matter decline. This is why processing speed shows the earliest and steepest age-related drops of any cognitive ability.

Crystallized intelligence, by contrast, relies on distributed semantic memory networks that are remarkably resilient to aging. Knowledge stored in long-term memory remains accessible far longer than the fluid processing systems that originally acquired it. The temporal lobes, hippocampus, and semantic networks that support vocabulary and factual knowledge show much more modest age-related changes than the frontal systems supporting fluid intelligence.

There is a genuine neurobiological basis for the common experience of getting slower but wiser. Your brain is not simply deteriorating. It is shifting the balance between systems, relying less on raw processing power and more on the accumulated knowledge structures that decades of experience have built.

The Seattle Longitudinal Study: Seven Decades of Watching Minds Age

The most ambitious study of cognitive aging ever conducted is the Seattle Longitudinal Study, initiated by K. Warner Schaie in 1956. Every seven years, researchers test the current participants and add new ones. Approximately 6,000 people have participated over its nearly seven-decade history, and remarkably, twenty-six individuals from the original sample were still being followed at last report.

The study’s findings have been consistently more optimistic than cross-sectional research would predict. Cross-sectional studies, which compare different people at different ages at a single point in time, tend to overestimate cognitive decline because they confuse cohort effects, such as differences in education and nutrition between generations, with actual aging effects. The Seattle study, which tracks the same individuals over time, provides a cleaner picture.

Its key findings include the demonstration that middle-aged adults in their forties and fifties actually perform better on four out of six cognitive tasks than those same individuals did when they were young adults. The abilities that showed improvement into middle age included verbal ability, spatial orientation, inductive reasoning, and verbal memory. Only perceptual speed and numerical ability showed consistent early declines.

Significant cognitive decline, the kind that meaningfully impairs daily functioning, was relatively rare before age sixty and did not become common until the mid-seventies. Even at age eighty-one, fewer than half of participants showed reliable decline over the previous seven years on any given ability.

The study also identified protective factors against cognitive decline: an intellectually active lifestyle, a flexible personality style, above-average education, marriage to a spouse with high cognitive ability, and maintained perceptual processing speed. These findings align with the cognitive reserve hypothesis I discussed in my previous article, the idea that intellectual engagement builds a buffer against age-related decline.

The Critical First Years: When IQ Is Most Volatile

If the story of aging is more nuanced than simple decline, the story of childhood cognitive development is more dynamic than most parents realize.

IQ scores in early childhood are notoriously unstable. A child tested at age three and retested at age ten may show swings of 10 to 20 points or more, not because of measurement error alone but because different cognitive abilities are developing at different rates and different children’s developmental curves have different shapes. Some children are cognitively precocious at four and average by twelve. Others are unremarkable at six and gifted by fourteen.

The Wilson Effect, documented by Ronald Wilson in the Louisville Twin Study and confirmed by subsequent research, demonstrates that the heritability of IQ increases dramatically from childhood to adulthood. At age five, genetic factors account for only about twenty percent of IQ variation, while shared family environment accounts for roughly fifty-five percent. By age eighteen, heritability has climbed to approximately eighty percent and shared environment has dropped to near zero.

This seemingly paradoxical finding reflects the increasing role of gene-environment correlation as children grow. Young children have their environments chosen for them. Adolescents and adults increasingly select environments that match their genetic predispositions, reading more if genetically inclined toward verbal ability, seeking out puzzles and strategic games if inclined toward spatial reasoning. Over time, the environment amplifies rather than counteracts genetic tendencies.

The practical implication for parents is important: an IQ score obtained at age four or five is a useful snapshot of current functioning but a poor predictor of adult intelligence. By age eight to ten, scores become substantially more stable, and by adolescence, they are reasonably predictive of adult cognitive ability, though still subject to change.

This is why I advise parents against making high-stakes educational decisions based on a single early-childhood IQ test. The developing brain is simply too dynamic, too responsive to environmental input, and too variable in its developmental timing to be reliably captured by a number at age four.

Adolescence: The Brain Under Construction

The adolescent brain deserves special attention because it represents a period of dramatic cognitive development that IQ tests capture only partially.

Between ages twelve and twenty-five, the brain undergoes extensive synaptic pruning and myelination, particularly in the prefrontal cortex. Unused neural connections are eliminated while frequently used pathways are strengthened and insulated with myelin, increasing their transmission speed. This process follows a back-to-front pattern: sensory and motor areas mature first, while the prefrontal cortex, responsible for planning, impulse control, and abstract reasoning, is among the last regions to complete development.

This has direct consequences for IQ test performance. Fluid intelligence measures, particularly those involving abstract reasoning and cognitive control, continue to improve through adolescence as prefrontal networks mature. Processing speed peaks in the late teens as myelination reaches its zenith. Working memory capacity, which depends on prefrontal function, may not reach its ceiling until the mid-twenties.

The timing of prefrontal maturation also explains something I encounter constantly in forensic evaluations of adolescents: the dissociation between measured intelligence and real-world judgment. A sixteen-year-old can score 120 on an IQ test, demonstrating excellent abstract reasoning in a calm, structured testing environment, and still make catastrophic decisions in emotionally charged real-world situations. The cognitive hardware is largely in place. The executive control systems that govern its deployment under conditions of emotional arousal, social pressure, and uncertainty are still under construction.

The Supreme Court has recognized this neuroscience. In Roper v. Simmons (2005), the Court cited developmental brain science in prohibiting the death penalty for crimes committed before age eighteen, and in Graham v. Florida (2010), it extended similar reasoning to sentences of life without parole for juvenile non-homicide offenders.

The Compensation Hypothesis: Can Wisdom Replace Speed?

One of the most compelling questions in cognitive aging research is whether gains in crystallized intelligence can functionally compensate for losses in fluid intelligence. The compensation hypothesis suggests that older adults can maintain real-world performance by increasingly relying on accumulated knowledge, practiced routines, and strategic approaches that bypass the need for raw processing speed.

The evidence is mixed but fascinating. Airline pilots provide a natural laboratory. Studies of pilot performance in flight simulators show that older pilots take longer to learn new simulator systems but outperform younger pilots at avoiding collisions. Their decades of pattern recognition compensate for their slower reaction times. Chess offers a similar picture: older masters may think more slowly than younger ones, but their vast library of stored game patterns allows them to evaluate positions with less exhaustive search.

However, a recent large-scale longitudinal analysis using data from both the Virginia Cognitive Aging Project and the Betula Project found a more sobering result: rates of change in fluid and crystallized abilities were strongly correlated at the individual level. People showing the greatest fluid declines also tended to show the smallest crystallized gains, or even crystallized losses. This suggests that some common factor, perhaps overall brain health or vascular integrity, drives decline across both systems simultaneously, limiting the scope for true compensation.

A 2025 paper published in Intelligence offered a more integrative perspective. Analyzing age-related trends across nine constructs associated with life success, including cognitive abilities, personality traits, emotional intelligence, financial literacy, moral reasoning, and cognitive flexibility, the researchers found that human functioning peaks in midlife, around ages fifty-five to sixty, when the composite of all these abilities reaches its maximum. The decline of fluid intelligence is offset by continued growth in emotional regulation, accumulated expertise, and social wisdom. Career success, not coincidentally, peaks in the same window.

What Cognitive Decline Actually Looks Like

One of the most common referrals I receive is from adults in their forties, fifties, and sixties who are worried that normal aging represents the beginning of dementia. Understanding the difference between normal cognitive aging and pathological decline is essential, both for reducing unnecessary anxiety and for catching genuine problems early.

Normal cognitive aging involves gradual, predictable changes: slower processing speed, reduced working memory capacity, more frequent word-finding difficulties (the “tip of the tongue” phenomenon), and some decline in the ability to learn and retain new information. These changes are real but modest in most people, and they are substantially offset by preserved or improved crystallized abilities.

What normal aging does not typically involve: getting lost in familiar environments, forgetting the names of close family members, losing the ability to manage finances or follow a conversation, personality changes, or progressive worsening that interferes with daily activities. These symptoms suggest pathological processes, potentially mild cognitive impairment or early dementia, and warrant immediate clinical evaluation.

A useful clinical rule of thumb: if you are aware of and bothered by your cognitive changes, that awareness itself is a reassuring sign. The most concerning cognitive declines are often the ones the person does not notice, because the very brain systems required for self-monitoring are among those affected.

The concept of cognitive reserve provides a framework for understanding why some people tolerate age-related brain changes better than others. Education, intellectual engagement, social connection, bilingualism, and occupational complexity all build reserves that allow the brain to sustain more damage before clinical symptoms emerge. A meta-analysis of over 437,000 subjects found that each year of education reduces dementia risk by approximately forty-six percent, not because education prevents brain pathology but because it provides alternative neural pathways and compensatory strategies that maintain function despite it.

Protecting Your Cognitive Future

The research on cognitive aging does not merely describe decline. It identifies modifiable factors that influence the rate, trajectory, and functional impact of age-related cognitive change.

Physical exercise has the strongest evidence base for preserving cognitive function in aging. As I discussed in my previous article, aerobic exercise increases hippocampal volume, upregulates BDNF, and produces consistent small-to-moderate cognitive benefits, particularly for executive function and processing speed. The effect is not about getting smarter. It is about slowing the rate of decline in the systems most vulnerable to aging.

Intellectual engagement, pursuing cognitively demanding activities throughout life, builds and maintains the cognitive reserve that buffers against decline. This does not mean brain training games, which, as we discussed, do not produce far transfer. It means genuine intellectual challenge: learning new skills, reading demanding material, engaging with complex problems, maintaining curiosity. The Seattle Longitudinal Study found that an intellectually active lifestyle was one of the strongest predictors of maintained cognitive function into old age.

Sleep becomes increasingly important with age, not less. Age-related changes in sleep architecture, including reduced deep sleep and more frequent nighttime awakenings, can compound the cognitive effects of biological aging. Chronic sleep restriction in older adults amplifies the very processing speed and working memory deficits that aging already produces.

Social connection protects cognition through mechanisms that are still being fully elucidated but likely involve cognitive stimulation, emotional regulation, and stress buffering. Loneliness and social isolation are now recognized as significant risk factors for cognitive decline and dementia.

And vascular health, perhaps surprisingly, is one of the strongest modifiable predictors of cognitive aging. The brain consumes approximately twenty percent of the body’s blood supply. Conditions that compromise vascular function, including hypertension, diabetes, high cholesterol, and smoking, directly impair the brain’s oxygen and nutrient supply. Managing cardiovascular risk factors in midlife is, in essence, managing cognitive risk factors.

A More Accurate Picture of Your Mind

Everything in this article points toward a fundamental reconceptualization of what intelligence means across the lifespan. The narrative of peak and decline, while containing a kernel of truth about fluid abilities, misses the richer, more complex, and ultimately more hopeful story that the science actually tells.

You are not simply “smarter” at twenty-two and “less smart” at sixty-two. You are a different kind of smart. The twenty-two-year-old brings raw processing power, rapid learning, and fluid reasoning to novel problems. The sixty-two-year-old brings four additional decades of accumulated knowledge, pattern recognition, emotional understanding, and practical wisdom. Both represent genuine cognitive strengths. Neither is globally superior.

This is why IQ tests are designed with age-normed scoring. Your score does not represent raw performance. It represents how you perform relative to other people your age. A sixty-five-year-old who scores 115 is not being graded on a curve to make her feel better. She is genuinely performing at the 84th percentile of cognitive ability for people with brains that have experienced sixty-five years of the same biological aging processes her brain has experienced.

Understanding your cognitive profile at your current age, knowing which abilities are at their peak, which are declining, and which are still growing, provides genuinely actionable information. It tells you when to trust your quick instincts and when to take more time. It tells you where your years of experience give you an advantage that no amount of raw processing speed can match. And it tells you where deliberate strategies and environmental supports can compensate for the areas where biology is working against you.

Final Thoughts

In two decades of assessing intelligence across the full span of human development, from three-year-olds just beginning to demonstrate abstract thought to ninety-year-olds navigating the cognitive challenges of extreme age, I have developed a deep respect for the brain’s capacity to adapt, compensate, and find new pathways to competence even as some of its original pathways narrow.

The woman I tested at twenty-two and again at forty-five was right that something had changed. But she was wrong about what it meant. She had not lost cognitive ability. She had traded one kind for another, exchanging some speed for substantial depth, some working memory bandwidth for vastly greater knowledge and judgment. By virtually every measure that matters for her professional life, her relationships, and her daily functioning, she was more cognitively capable at forty-five than she had been at twenty-two.

The science of cognitive aging tells us that your mind is not a candle burning down. It is a landscape being reshaped, some hills eroding while new ones rise, the overall terrain becoming more complex and, in many ways, more useful with each passing decade.

Your IQ score captures a snapshot of that landscape at a single moment. Understanding how the terrain shifts over time, and what you can do to maintain its most valuable features, is the beginning of aging wisely.