Can Fluid Intelligence Be Trained? What Transfers—and What Stays Specific to Practice

You can become dramatically better at a demanding brain task. The harder question is whether that improvement follows you into unfamiliar problems you never practiced.

Fluid intelligence can change, but short cognitive-training programs have not shown a reliable way to produce large, broad increases in it. Training usually improves the practiced task and sometimes closely related tasks. Evidence becomes much weaker when researchers ask whether those gains transfer to genuinely different measures of fluid intelligence.

That distinction—between getting better at a task and changing a broad cognitive ability—is the heart of the debate.

What Is Fluid Intelligence?

Fluid intelligence is the ability to reason through new problems without depending heavily on previously learned facts. It helps when you have to discover a rule, compare relationships, reorganize information, or solve something unfamiliar.

Matrix problems are a classic example. You see a pattern of shapes with one piece missing and infer the rule that completes it. Our guide to fluid vs. crystallized intelligence explains how this kind of novel reasoning differs from accumulated knowledge.

The appeal of training fluid intelligence is obvious. If you could strengthen a broad reasoning capacity, improvement might carry into many unrelated tasks. Researchers call that kind of generalization far transfer.

Why Working-Memory Training Looked So Promising

Working memory and fluid intelligence are strongly related, so researchers wondered whether training one could improve the other.

A famous 2008 study by Jaeggi and colleagues reported that adults who trained on a demanding working-memory task subsequently improved on measures of fluid intelligence. The authors also found a relationship between the amount of training and the size of the fluid-intelligence gain.

That finding attracted enormous attention because the trained task differed from the intelligence tests used to measure transfer. It looked like evidence that practice had escaped the boundaries of the exercise itself.

The result also helped make N-Back training one of the best-known approaches in the cognitive-training world.

Then the Replications Became Complicated

Later studies produced a much less tidy picture.

A 2015 meta-analysis focused specifically on N-Back training. Across 20 qualifying studies of healthy adults, the researchers found a small positive effect on fluid-intelligence measures. So the early result did not simply disappear.

The important question became how robust that effect was. Cognitive-training studies are unusually sensitive to the choice of control group. If the training group spends weeks doing an engaging task while the control group does nothing, improvements can reflect expectations, repeated testing, motivation, or other nonspecific effects.

A broader 2016 meta-analysis examined 87 publications containing 145 experimental comparisons. Working-memory training reliably improved measures close to the trained memory skills. When the comparison used treated control groups, the authors found no convincing evidence of reliable far transfer to nonverbal ability, verbal ability, reading, or arithmetic.

A Two-Year Training Study Put the Idea Under Pressure

One objection to short training studies is obvious: perhaps a few weeks is simply too little time to change a broad ability. A 2022 study tested that possibility with an unusually long intervention.

Students practiced a varied set of working-memory tasks over two years. The training produced substantial, reliable improvement on the practiced tasks and on a latent working-memory factor built from them. Yet researchers found no transfer to fluid or crystallized intelligence compared with the control group.

That result is especially informative because training clearly worked on its intended target. The problem was generalization.

Does This Mean Fluid Intelligence Is Fixed?

No. Evidence against broad transfer from a particular training method does not establish that fluid intelligence can never change.

Measured intelligence changes across development, and sustained education can produce genuine cognitive gains. A large meta-analysis using quasi-experimental evidence found that an additional year of education was associated with roughly 1 to 5 IQ points of improvement across the studies analyzed.

Our article on whether you can increase your IQ looks at those broader sources of change. The key lesson here is narrower: success in a cognitive-training exercise is not automatically evidence that fluid intelligence itself increased.

What About Practicing Matrix Reasoning Directly?

If you practice matrix problems, you should expect to become more comfortable with matrix problems. You may recognize common transformations faster, organize the visual information more efficiently, and make fewer avoidable mistakes.

Those are useful gains. They can also make it difficult to tell how much of a higher matrix score reflects broader reasoning ability and how much reflects familiarity with the task family.

This is why researchers use multiple measures when they want strong evidence of far transfer. A gain becomes more persuasive when it appears on unfamiliar tasks that share the underlying ability without sharing the same surface rules.

For the history and logic of this format, see our guide to Raven’s Progressive Matrices.

Try a Fluid-Reasoning-Style Task

Matrix reasoning is one way to experience the kind of novel rule discovery often used to sample fluid reasoning. Use the test below as practice and as a snapshot of your current performance. Improvement on repeated matrix problems is meaningful practice progress; it should not be treated by itself as proof of a broad increase in fluid intelligence.

So What Is Worth Training?

Train a skill when you genuinely want that skill. Working-memory exercises can improve working-memory performance. Matrix practice can sharpen your approach to visual pattern problems. Strategy learning can make you more efficient at particular forms of reasoning.

For broader cognitive development, sustained learning has stronger evidence than a short “brain boost” routine. Cognitive Train’s guide to how to increase IQ separates those longer-term approaches from methods that mainly improve a practiced task.

The same transfer issue is central to the question of whether brain games make you smarter and whether you can train your brain to be smarter.

The Best Answer Is More Precise Than Yes or No

Fluid-intelligence test performance can improve, and some studies have reported transfer after cognitive training. The broader literature gives much less confidence that short, targeted training reliably raises fluid intelligence across genuinely different tasks.

So if you get better at a demanding reasoning exercise, count that improvement for what it is. Then ask the harder question: does the gain survive a new task, a new format, and a new set of rules? That is where evidence for a broader cognitive change begins.

Explore Cognitive Train’s brain tests and cognitive assessments, or browse our cognitive training and brain training tools for more ways to practice reasoning, memory, attention, and processing speed.