Working memory is the cognitive system most closely linked to fluid intelligence - and it is partially trainable, though not as dramatically as some claims suggest. Here are 10 methods ranked by evidence quality and practical effect size.
Tier 1: Strong Evidence, Large Effects
1. Sleep (Effect: Massive)
Sleep deprivation devastates working memory capacity before affecting almost any other cognitive function. A single night of 6 hours reduces working memory span by 1โ2 items and dramatically increases error rates on complex span tasks. During slow-wave sleep, the hippocampus replays and consolidates newly encoded information. This consolidation is not optional for working memory maintenance - it is the biological mechanism by which today's learning becomes tomorrow's accessible knowledge.
2. Aerobic Exercise (Effect: +5โ15% capacity)
A 2018 meta-analysis of 80 randomized controlled trials found that acute aerobic exercise (20โ30 minutes) significantly improved working memory performance, with the largest gains in tasks requiring active manipulation rather than passive storage. The mechanism: exercise increases BDNF (brain-derived neurotrophic factor), which promotes hippocampal neuroplasticity. Long-term regular cardio is associated with increased hippocampal volume - the brain structure most central to memory.
Tier 2: Good Evidence, Moderate Effects
3. N-Back Training (Effect: +10โ20% on trained tasks)
N-Back training reliably improves performance on N-Back tests and closely related measures. Transfer to broad fluid intelligence is weak and inconsistent across studies. The practical takeaway: N-Back is a genuine working memory workout, not a general cognitive enhancer. Regular N-Back practice on ReflexBenchmark will improve your N-Back score and likely help with tasks that specifically require the type of updating it trains. Take our N-Back test as a baseline.
4. Chunking (Effect: +30โ100% apparent capacity)
Chunking does not increase raw working memory capacity but dramatically increases effective capacity by compressing information into fewer, larger units. A phone number (9821045678) is 10 separate items - at the edge of most people's span. Chunked into three units (982-1045-678), it is 3 items - well within capacity. Expert chess players can hold far more board positions in working memory than novices - not because their capacity is larger, but because they chunk positions into meaningful patterns.
5. Reduce Multitasking (Effect: Preserve existing capacity)
Habitual media multitasking is associated with reduced working memory capacity and worse performance on tasks requiring sustained attention. This is likely a bidirectional relationship - people with weaker working memory may be more prone to multitasking - but reducing habitual task-switching is consistently associated with better working memory performance.
Tier 3: Moderate Evidence, Smaller Effects
6. Mindfulness Meditation (Effect: +5โ10% after 8 weeks)
Several RCTs have found that 8-week mindfulness-based programs improve working memory capacity, particularly in high-stress populations where cognitive resources are chronically depleted by rumination and anxiety. The mechanism: mindfulness reduces mind-wandering, freeing working memory resources that would otherwise be occupied with off-task thoughts.
7. Video Games - Action Genre (Effect: +5โ10%)
Multiple studies (Bavelier et al., Green et al.) have found that action video game players show working memory advantages compared to non-players. Action games require tracking multiple moving objects, updating priorities rapidly, and ignoring distractors - directly training working memory functions. Fast-paced games with demanding visual environments show the strongest associations.
8. Optimize Your Environment (Effect: Variable)
Working memory operates in limited capacity - distractions consume resources. A phone on the desk (even face down and silent) measurably reduces working memory performance in some studies, presumably through attentional capture. Noise-canceling headphones, phone-free work sessions, and single-task focus all preserve working memory capacity for the task at hand.
Tier 4: Weak or Inconsistent Evidence
9. Omega-3 Supplementation
Some studies show modest working memory benefits from omega-3 supplementation (particularly DHA), especially in populations with low dietary intake. Effects in adults with adequate dietary omega-3 are inconsistent. Whole food sources (fatty fish, walnuts) show better associations than supplements in observational data.
10. Dual N-Back (Effect: Debated)
The dual N-Back variant - tracking both position and sound simultaneously - generated enormous interest after the Jaeggi et al. (2008) study claiming fluid intelligence transfer. Subsequent replications have been mixed. Current position: dual N-Back reliably trains the trained task; broader intelligence transfer is not reliably replicated. Still a good working memory workout, just not a magic IQ pill.
๐ง Benchmark Your Working Memory
Take the N-Back test to measure your current working memory updating capacity.
Take N-Back Test โFrequently Asked Questions
This is genuinely debated in research. A 2015 meta-analysis of 20 studies found a small but significant positive effect of N-Back training on fluid intelligence, but other reviews and re-analyses have raised concerns about study design and question whether the effect reflects real transfer or motivational differences between groups.
They're related but distinct - short-term memory refers to briefly holding information, while working memory involves actively manipulating that information, such as reordering, calculating, or comparing it while it's held in mind.
Near transfer refers to improvement on tasks very similar to the trained task, which is well supported by research. Far transfer means improvement on unrelated cognitive abilities like general intelligence, which is much more contested.
Chunking, or grouping information into meaningful units, can substantially increase apparent capacity - for example, remembering a 10-digit number as chunked groups is far easier than as 10 separate digits, even though raw storage capacity hasn't changed.
Working memory is described as a bottleneck because most higher-order cognitive tasks, like solving multi-step problems, require holding several pieces of information in mind simultaneously - limited capacity here constrains performance on those tasks.
Some research suggests far-transfer effects to fluid intelligence may be more consistently found in older adults compared to younger, healthy adults, though results across studies remain mixed.
Evidence for Omega-3 supplementation specifically improving working memory is weak and inconsistent compared to well-supported interventions like sleep and aerobic exercise, so it shouldn't be relied on as a primary strategy.
Sleep deprivation is one of the most consistently documented factors that impairs working memory performance, with effects that can be large enough to meaningfully affect both trained and untrained memory tasks.
Heavy multitasking doesn't appear to cause permanent damage, but it can interfere with encoding information effectively in the moment, and reducing multitasking is associated with better preservation of existing working memory performance.
Yes, some caution is warranted - while working memory tasks can improve performance on similar tasks, the research on broad intelligence gains is contested, and strong marketing claims often outpace what the evidence actually supports.