Working Memory Exercises for Students and Professionals

Dual n-back requires simultaneously tracking visual position and auditory identity n steps back in a continuous stream. This double-updating demand strongly exercises the central executive — the component with the broadest far-transfer effects. Jaeggi et al. (2008) found 19 days of dual n-back training produced gains in fluid intelligence and working memory on untrained tasks. Subsequent replications show consistent but smaller effects (d≈0.3–0.5 on far-transfer tasks). Best for: students in math-heavy disciplines, software developers, data analysts.

The reading-span task requires reading a sentence, judging its meaning, and simultaneously memorizing the final word of each sentence — then recalling all the final words. This complex span task directly mimics the demands of reading comprehension and has shown the most reliable far-transfer to academic reading performance in students. Multiple studies show 4–6 weeks of adaptive reading-span training improves reading comprehension scores by 15–25%.

Deliberately practice reading syntactically complex text (embedded relative clauses, passive voice, center-embedded structures) at the edge of your comprehension. Do not simplify — complexity is the training stimulus. The working memory demands of parsing difficult text are real and trainable. 15 minutes of complex reading daily produces measurable WM improvements within 3 weeks for most adults. Good sources: academic papers in your field, complex literature, philosophy texts.

Using the Sequence Memory and Number Memory tests as training tools — not just assessments — provides the adaptive difficulty that produces the best WM training outcomes. The key is deliberate practice at ceiling level (failing regularly), not comfortable practice within your current range. The measurability of your score makes progress tracking objective.

Mental arithmetic directly exercises the phonological loop and central executive in tandem. Practice multi-step problems at the edge of difficulty (3-digit × 2-digit multiplication, multi-step word problems) without writing anything down. This transfers specifically to math performance and number-handling tasks in professional contexts. Progress along this axis can be tracked with the Number Memory test, which taps the same phonological loop.

Read or listen to content, then wait 15 minutes before writing a summary from memory. The delay forces working memory to actively maintain and organize information rather than copying it immediately. This is essentially a working memory training task that simultaneously improves comprehension, retention, and written communication skills. Particularly effective for students preparing for essay exams or professionals who must brief others after meetings.

Producing speech in a second language — not passive vocabulary review — is one of the most demanding real-world working memory exercises. You must hold semantic intent, search for vocabulary, maintain sentence structure, and suppress L1 interference simultaneously. Adults learning a new language show measurably improved executive function and working memory relative to non-learners when practice involves active production (speaking, writing) rather than passive input.

Lumosity, BrainHQ, Elevate, and similar apps produce strong improvements on their own tasks but show weak or inconsistent far-transfer in the majority of rigorous RCTs. The FTC's 2016 settlement with Lumosity — citing deceptive advertising about cognitive benefits — reflects the scientific consensus. These apps are not harmful, but they should not be the core of a WM training program. See the full analysis in our article on the brain training myth.