Does Dual N-Back Training Increase IQ?

Written and researched by the Studio âge web research team / Last reviewed: / Editorial policy

Dual n-back training and intelligence research

Quick answer

Dual n-back training may improve scores on some fluid-intelligence tests, but it should not be described as a proven way to produce large, permanent IQ gains. The most defensible summary is nuanced: dual n-back practice reliably improves the task being trained and often improves closely related working-memory tasks. Evidence for broader transfer to fluid intelligence is mixed. Some meta-analyses report a small positive effect; other reviews argue that much of the benefit is task-specific and that far transfer to general intelligence is weak or uncertain.

In practical terms, dual n-back is a demanding working-memory exercise. It is reasonable to use it as a focused attention and working-memory training task. It is not reasonable to promise that a few weeks of practice will make every user smarter in a broad, real-world sense. The scientific question is not simply "yes" or "no." The better question is: what transfers, how far does it transfer, and under what conditions?

What fluid intelligence is and why it matters

Fluid intelligence, often abbreviated as Gf, refers to the ability to reason through novel problems without relying mainly on previously learned knowledge. It is involved when a person detects patterns, solves unfamiliar puzzles, adapts to new rules, or works through abstract reasoning tasks. Fluid intelligence is related to, but not identical with, IQ. IQ tests usually include multiple components, such as vocabulary, processing speed, working memory, and reasoning. Fluid-intelligence tests focus more narrowly on novel reasoning.

The distinction matters because many dual n-back studies do not measure "IQ" in the broad everyday sense. They measure performance on laboratory tests of Gf, such as matrix reasoning tasks. When a study reports improved fluid intelligence, that does not automatically mean a person will show broad gains in school, work, decision-making, or every part of an IQ battery. It means performance improved on a particular class of reasoning measures.

This is why precise wording matters. A careful statement is: "Some studies have reported gains on fluid-intelligence measures after n-back training." A stronger claim, such as "dual n-back increases IQ," requires more evidence, especially evidence that improvements generalize beyond similar cognitive tests and persist over time.

The original claim: Jaeggi et al. (2008)

The modern debate began with a highly influential paper: Jaeggi, Buschkuehl, Jonides, and Perrig (2008), "Improving fluid intelligence with training on working memory," Proceedings of the National Academy of Sciences, 105(19), 6829-6833, DOI: 10.1073/pnas.0801268105, PMID: 18443283. Jaeggi et al. (2008) used an adaptive dual n-back task as the training intervention. Participants had to monitor visual and auditory streams at the same time, compare each current stimulus with the one presented n trials earlier, and respond when a match occurred.

Jaeggi et al. (2008) reported that working-memory training transferred to measures of fluid intelligence. The authors also reported a dose-response pattern: more training was associated with larger gains on Gf tests. The claim was important because the participants did not simply practice the same test used to measure improvement. The training task and the outcome test were different, so the study appeared to show far transfer from working-memory training to abstract reasoning.

The 2008 result attracted attention because it challenged a long-standing concern in cognitive training: people often improve at the exact task they practice, but those improvements do not necessarily generalize. If dual n-back could improve fluid intelligence, it would be a rare example of a short cognitive-training program affecting a broader mental ability. That possibility made the study influential, but it also made the claim a target for replication attempts and methodological criticism.

The skeptical response: Au et al. (2015) and later meta-analysis

A key pro-transfer synthesis is Au, Sheehan, Tsai, Duncan, Buschkuehl, and Jaeggi (2015), "Improving fluid intelligence with training on working memory: A meta-analysis," Psychonomic Bulletin & Review, 22(2), 366-377, DOI: 10.3758/s13423-014-0699-x, PMID: 25102926. Au et al. (2015) focused on n-back training studies that included fluid-intelligence outcomes, control groups, and healthy adult participants. The authors included 20 studies and concluded that n-back training produced a small but statistically significant positive effect on Gf.

That finding supports a cautious "maybe" answer. It does not show that dual n-back produces large IQ gains. It suggests that short-term n-back training can improve performance on laboratory measures of fluid intelligence by a modest amount under some conditions. The size and reliability of that effect remain important. A small average effect may be meaningful for theory, but it is different from a large, predictable benefit for every user.

There is also an important citation correction. Some summaries attach PMID: 28116702 to Au et al. (2015), but that PubMed ID refers to another important paper: Soveri, Antfolk, Karlsson, Salo, and Laine (2017), "Working memory training revisited: A multi-level meta-analysis of n-back training studies," Psychonomic Bulletin & Review, 24(4), 1077-1096, DOI: 10.3758/s13423-016-1217-0, PMID: 28116702. Soveri et al. (2017) analyzed 33 randomized controlled trials and 203 effect sizes. They found a medium-sized transfer effect to untrained n-back tasks, but very small effects for other working-memory tasks, fluid intelligence, and cognitive control. They concluded that a substantial part of transfer after n-back training is task-specific.

The counter-argument: near transfer vs. far transfer

The near-transfer versus far-transfer distinction is the center of the debate. Near transfer means that training improves performance on tasks that are very similar to the training task. For dual n-back, near transfer might include another n-back version with different stimuli, a similar working-memory updating task, or a closely related attention task. Near transfer is plausible because the trained and transfer tasks share processes such as updating, monitoring, and response control.

Far transfer means that training improves a more distant ability, such as broad IQ, academic achievement, job performance, or general reasoning. Far transfer is harder to establish because many alternative explanations must be ruled out. A person might improve because of test familiarity, motivation, expectancy effects, strategy learning, regression to the mean, or differences between active and passive control groups. Strong far-transfer evidence requires well-designed randomized studies, active control groups, preregistered outcomes, adequate sample sizes, and follow-up testing.

The skeptical position is not that the brain cannot learn. It is that learning is often specific. Dual n-back clearly trains a person to handle dual n-back demands. The uncertain part is whether that training changes a general cognitive capacity in a way that carries over to unrelated tasks. This is why researchers often avoid saying "n-back increases IQ" and instead say "n-back training may produce modest transfer to fluid-intelligence measures, but the extent of far transfer is debated."

What the evidence reviewed through 2024 supports

Based on the evidence reviewed here through 2024, the most balanced conclusion is: n-back training improves performance on the trained task and can transfer to closely related tasks; evidence for broad, durable IQ gains is limited and contested. Meta-analyses differ partly because they include different studies, use different statistical models, and classify transfer outcomes differently. Some analyses find small positive effects on Gf. Others emphasize that the most reliable effects are task-specific and that far-transfer effects are small or unstable.

A useful recent synthesis is Syed, Lum, Byrne, and Skvarc (2024), "Examining Working Memory Training for Healthy Adults-A Second-Order Meta-Analysis," Journal of Intelligence, 12(11), 114, DOI: 10.3390/jintelligence12110114. This second-order meta-analysis aggregated evidence from meta-analyses of working-memory interventions in healthy adults. It concluded that working memory can improve through training, but the average effect is small and real-life cognitive utility may be minimal. This conclusion fits the broader pattern: training effects are real enough to study, but not strong enough to justify exaggerated claims.

Therefore, the current evidence supports a conservative recommendation. Dual n-back is worth trying if you want a challenging working-memory exercise, enjoy measurable practice, and understand that the main expected gains are task performance, attention control, and possibly small improvements on related laboratory measures. It should not be marketed as a guaranteed IQ-increase intervention.

Bottom line

Does dual n-back training increase IQ? The best evidence-based answer is: not in the strong popular sense. Jaeggi et al. (2008) reported fluid-intelligence gains after adaptive dual n-back training. Au et al. (2015) found a small positive meta-analytic effect across 20 studies. Soveri et al. (2017, PMID: 28116702) found that transfer is strongest for untrained n-back tasks and very small for fluid intelligence and other domains. The evidence reviewed through 2024 supports treating dual n-back as a legitimate working-memory training task with debated and likely modest far-transfer effects.

If you try dual n-back, treat it like cognitive practice rather than a miracle intervention. Start with a manageable n-level, train while focused, track accuracy, and increase difficulty only when performance is stable. The value of the task is clearest when used as a structured way to practice working memory and sustained attention.

Try the free browser N-Back tool

References