Does Dual N-Back Training Help ADHD? What the Research Says

What Is the Connection Between ADHD and Working Memory?

Working memory is the mental workspace that holds information in mind while you act on it — tracking the beginning of a sentence while reading the end, holding a phone number while you walk to write it down, or managing multiple steps of a task without losing your place. For most adults with ADHD, this system is measurably impaired.

Research puts the prevalence of working memory deficits in ADHD at 75–85% of individuals when assessed with high-demand tasks (Kofler et al., 2024, Nature Reviews Psychology). One meta-regression found that up to 98% of people with ADHD score below average on construct-valid working memory tests. These are not small gaps — meta-analyses report effect sizes of Cohen's d = 1.06 to 2.15 for spatial working memory, meaning the average person with ADHD performs roughly one to two standard deviations below neurotypical controls.

For adults specifically, a 2013 meta-analysis by Alderson et al. — the first to analyze adults exclusively — reviewed 38 studies and found consistent moderate deficits in both phonological working memory (d = 0.55) and visuospatial working memory (d = 0.49). Critically, these deficits persist across the lifespan, though they appear to decrease slightly with age.

In daily life, impaired working memory in ADHD often shows up as:

• Losing track of conversations because you can't hold both the thread and your response simultaneously
• Starting multi-step tasks and forgetting intermediate steps
• Missing deadlines and appointments even when you care about them
• Difficulty with mental math, following complex verbal instructions, or managing finances
• Starting sentences and losing your train of thought mid-way

Kofler et al. (2018) found that working memory deficits of this magnitude (d = 1.24) predicted real-world organizational problems (d = 0.85), with 38–57% of working memory's effect on organization flowing through inattentive behavior. In other words, the cognitive impairment you can measure in the lab maps directly onto the difficulties you experience at work.

Is ADHD fundamentally a working memory disorder?

Three major theoretical models — Barkley's Executive Function Model, Rapport's Working Memory Model, and Castellanos & Tannock's Endophenotype Framework — position working memory at or near the center of ADHD. Barkley's updated model (2022) elevated working memory from a mediating variable to a primary driver of ADHD symptoms, alongside behavioral inhibition, with better statistical fit than the original 1997 version. Rapport's model goes further, proposing that inattention, hyperactivity, and impulsivity are downstream consequences of a core working memory deficit — and that experimentally manipulating working memory demands can directly provoke or reduce ADHD behavioral symptoms in the lab.

This theoretical framing makes working memory an attractive training target: if you can strengthen the core deficit, the downstream symptoms should follow — though as the research below shows, this transfer has proven difficult to demonstrate in practice.

What Does the Research Say? Dual N-Back Studies in ADHD

Dual n-back is a working memory exercise where you simultaneously track a sequence of grid positions and spoken letters, responding each time the current stimulus matches what appeared n trials back. It places near-maximum demands on the central executive — the updating, manipulation, and serial-ordering components of working memory that are most impaired in ADHD.

Four peer-reviewed studies have tested dual n-back training directly in adults with ADHD.

Study 1: Attention Networks in ADHD Adults (Mesrobian et al., 2020)

Published in: Brain Sciences (PMC7600375)  |  Sample: 106 young adults (non-medicated ADHD, medicated ADHD, and matched controls)

Participants completed 20 blocks of adaptive dual n-back daily (~30 minutes/day) for 20 days within a single month. The study used a direct comparison of adaptive training versus a fixed difficulty 1-back control condition.

Key findings: Adaptive training improved the Conflict Effect — the executive control component of the Attention Network Task — in ADHD patients regardless of medication status, but not in neurotypical controls. Alerting and orienting networks were unaffected. The authors concluded: adaptive dual n-back "offers a promising candidate for cognitive remediation of adult ADHD patients without pharmaceutical medication."

The authors note this may suggest that people with greater working memory impairment have the most to gain from training, though they acknowledge this remains speculative.

Study 2: Boosting Working Memory in ADHD (Lintas et al., 2025)

Published in: Brain Sciences, 15(9):998  |  Sample: 106 participants (33 non-medicated ADHD, 42 medicated ADHD, 45 controls)

A follow-up study by the same research group reported improvements on the WAIS-IV Working Memory Index — a standardized, widely-used clinical test of working memory capacity. For Digit Span Backward (a task requiring mental manipulation of remembered sequences), the adaptive training group showed an effect size of Cohen's d = 0.46 versus d = 0.27 for fixed training. Transfer to visuospatial working memory (Corsi Block-Tapping) was modest.

Study 3: Training Restores Aberrant Brain Activity (Salmi et al., 2020)

Published in: Human Brain Mapping  |  Design: Randomized controlled trial

44 adults with ADHD were randomized to adaptive dual n-back or an active control (Bejeweled 2) for 5 weeks, 3 sessions per week. fMRI was conducted before and after training.

Key findings: Training improved both trained WM tasks and closely related measures (digit span, visuospatial running memory). More strikingly, fMRI data showed redistribution of working memory-related brain activity, most prominent in the prefrontal cortex — the region most consistently underactive in ADHD. The authors described it as partial restoration of aberrant ADHD brain activity, providing the first evidence that adult ADHD brain function remains responsive to working memory training.

Study 4: Functional Connectivity RCT (Tolonen et al., 2024)

Published in: NeuroImage: Clinical  |  Design: Randomized controlled trial with neuroimaging

42 adults with ADHD and 36 neurotypical controls were studied. ADHD participants showed decreased functional connectivity during n-back tasks across fronto-parietal, temporal, occipital, cerebellar, and subcortical networks. The training group showed WM-related connectivity changes associated with improved performance on the Continuous Performance Test.

What these four studies show

Across all four studies, adaptive dual n-back training in ADHD adults reliably produces:

• Improved performance on trained WM tasks
• Near-transfer to closely related WM tests (verbal > visuospatial)
• Measurable changes in brain activity patterns associated with ADHD

The Harder Question: Does Working Memory Improvement Reduce ADHD Symptoms?

This is where the research becomes genuinely complicated, and where honest reporting matters.

What the meta-analyses find

Seven major meta-analyses have examined cognitive training for ADHD, and they converge on a finding that depends heavily on who is assessing the outcomes.

When assessed by unblinded observers (parents, teachers, or users who know whether they received the real training or a placebo):

• Sonuga-Barke et al. (2013): SMD = 0.64 for ADHD symptoms
• Cortese et al. (2015): SMD = 0.37 for total ADHD symptoms
• Spencer-Smith & Klingberg (2015): SMD = −0.47 for inattention

These look promising. The problem is expectancy effects. When participants know they're receiving "brain training," they tend to rate themselves better regardless of actual cognitive change.

When assessed by probably blinded observers (who don't know which condition the participant was in):

The 2023 Westwood meta-analysis — the most rigorous to date, covering 36 RCTs and published in Molecular Psychiatry — concluded: "No empirical support for the use of computerized cognitive training as a stand-alone intervention for ADHD symptoms."

But working memory itself does improve

Crucially, the same meta-analyses consistently find robust improvements in working memory performance: verbal WM (SMD ≈ 0.38–0.52) and visuospatial WM (SMD ≈ 0.47–0.66). These effects survive blinding. What doesn't reliably transfer are reductions in clinical ADHD symptom scores on rating scales.

Rapport et al. (2013) identified a possible reason: most working memory training targets simple storage and rehearsal, but ADHD's core deficit involves the central executive — updating, manipulation, and serial reordering — which requires much higher cognitive demands. Dual n-back is specifically designed for this; simpler span tasks are not.

How does dual n-back compare to other ADHD interventions?

Stimulant medication is approximately 4–8× the effect size of cognitive training for ADHD symptom reduction. Neurofeedback — often marketed aggressively to ADHD patients at $3,000–$10,000+ for a course — shows no significant effect on ADHD symptoms when properly blinded (JAMA Psychiatry, Westwood et al., 2025).

What professional bodies say

The American Academy of Pediatrics (Wolraich et al., 2019) lists cognitive training among interventions with "too little evidence to recommend" and explicitly does not include it in treatment guidelines. The UK's NICE (NG87, 2018) recommends parent training, medication, and CBT, with no mention of cognitive training. CHADD takes a cautious position: cognitive training may improve certain lab-based cognitive skills, but "effects are small and the skills that are improved are generally lab-based skills."

So Should Adults With ADHD Try Dual N-Back?

The honest answer: it depends on what you're trying to achieve.

If you're looking for working memory improvement — reduced forgetfulness, better multitasking, stronger focus when holding information in mind — the evidence supports dual n-back as a well-designed training tool. The near-transfer effects are reliable, adaptive training is clearly superior to fixed difficulty, and there is preliminary evidence of genuine neurological change in ADHD brains.

If you're hoping dual n-back will replace or meaningfully substitute for medication or therapy, the evidence does not support that expectation. Blinded assessments consistently show small to non-significant effects on core ADHD symptom scores. Medication remains the most effective pharmacological option; CBT the most effective non-pharmacological one.

The most rational framing, consistent with the research: dual n-back as a complementary tool. Something you do alongside whatever treatment plan you're already following, with the realistic expectation of improving working memory capacity specifically — not as a cure for ADHD.

Practical considerations for getting the most from training:

• Use adaptive difficulty — every study finding benefits in ADHD used adaptive mode. Fixed difficulty training consistently underperforms.
• Train consistently: Studies used 18–25 sessions over 4–5 weeks, at 25–45 minutes per session. Daily or near-daily practice outperforms sporadic longer sessions.
• Track d-prime, not just accuracy — raw accuracy is misleading. A 90% score can mean excellent performance or lucky guessing. D-prime (the signal detection measure from cognitive research) separates true working memory sensitivity from noise.
• Give it at least 4 weeks before evaluating results. The fMRI changes in the Salmi et al. (2020) study required 5 weeks of consistent training to appear.

Frequently Asked Questions

The Bottom Line

The research supports a nuanced picture: dual n-back training reliably strengthens working memory in adults with ADHD, and early evidence suggests it may partially restore the atypical brain activity patterns associated with the condition. The question of whether this translates to symptom relief in daily life remains genuinely open — blinded assessments show small effects that are often non-significant on clinical rating scales, and no major guideline currently recommends cognitive training as a treatment.

What this means practically: dual n-back is a well-evidenced tool for improving working memory specifically. Adults with ADHD who struggle most with WM-heavy tasks — tracking conversations, managing multi-step work, holding information in mind — have the clearest potential to benefit. It should not replace medication or therapy, but the evidence base is meaningfully better than alternatives like neurofeedback, and the cost of trying it is essentially zero.

Sources

1. [1] Kofler, M.J. et al. (2024). Executive function deficits in ADHD and ASD. Nature Reviews Psychology, 3(10), 701–719.
2. [2] Alderson, R.M. et al. (2013). ADHD and working memory in adults: A meta-analytic review. Neuropsychology, 27(3), 287–302.
3. [3] Kofler, M.J. et al. (2018). Working memory and organizational skills problems in ADHD. Journal of the American Academy of Child & Adolescent Psychiatry.
4. [4] Barkley, R.A. (1997). Behavioral inhibition, sustained attention, and executive functions. Psychological Bulletin, 121(1), 65–94.
5. [5] Castellanos, F.X. & Tannock, R. (2002). Neuroscience of ADHD: The search for endophenotypes. Nature Reviews Neuroscience, 3(8), 617–628.
6. [6] Willcutt, E.G. et al. (2005). Validity of the executive function theory of ADHD. Biological Psychiatry, 57(11), 1336–1346.
7. [7] Mesrobian, S.K. et al. (2020). Attention Networks in ADHD Adults after Working Memory Training with a Dual N-Back Task. Brain Sciences, PMC7600375.
8. [8] Lintas, A., Bader, M. & Villa, A.E.P. (2025). Boosting Working Memory in ADHD: Adaptive Dual N-Back Training Enhances WAIS-IV Performance. Brain Sciences, 15(9):998.
9. [9] Salmi, J., Soveri, A., Salmela, V. et al. (2020). Working memory training restores aberrant brain activity in adult ADHD. Human Brain Mapping, 41(17), 4876–4891.
10. [10] Tolonen, T., Leppamäki, S., Roine, T. et al. (2024). Working memory related functional connectivity in adult ADHD and its amenability to training. NeuroImage: Clinical, 44, 103696.
11. [11] Sonuga-Barke, E.J.S. et al. (2013). Nonpharmacological interventions for ADHD. American Journal of Psychiatry, 170(3), 275–289.
12. [12] Cortese, S. et al. (2015). Cognitive training for ADHD: Meta-analysis. JAACAP, 54(3), 164–174.
13. [13] Westwood, S.J. et al. (2023). Computerized cognitive training in ADHD: Meta-analysis with blinded outcomes. Molecular Psychiatry, 28, 1402–1414.
14. [14] Rapport, M.D. et al. (2013). Do programs designed to train working memory… work? Clinical Psychology Review, 33(8), 1237–1252.
15. [15] Melby-Lervåg, M., Redick, T.S. & Hulme, C. (2016). Working memory training does not improve performance on measures of intelligence or other measures of "far transfer." Perspectives on Psychological Science, 11(4), 512–534.
16. [16] Spencer-Smith, M. & Klingberg, T. (2015). Benefits of a working memory training program for inattention. PLoS ONE, 10(3), e0119522.
17. [17] Chacko, A. et al. (2014). A randomized clinical trial of Cogmed Working Memory Training in ADHD. Journal of Child Psychology and Psychiatry, 55(3), 247–255.
18. [18] Wolraich, M.L. et al. (2019). AAP Clinical Practice Guideline for ADHD. Pediatrics, 144(4), e20192528.
19. [19] NICE Guideline NG87 (2018). Attention deficit hyperactivity disorder: diagnosis and management.
20. [20] Westwood, S.J. et al. (2025). Neurofeedback for Attention-Deficit/Hyperactivity Disorder: A Systematic Review and Meta-Analysis. JAMA Psychiatry, 82(2), 118–129.
21. [21] McNab, F. et al. (2009). Changes in cortical dopamine D1 receptor binding associated with cognitive training. Science.