What does research say about multitasking vs single-tasking?

Why most multitasking advice is built on one oversimplified statistic

Most productivity writing on multitasking leans on a single, endlessly recycled claim: that multitasking cuts efficiency by some dramatic percentage. The problem is that "multitasking" is not one thing. Psychologists have spent decades separating it into distinct phenomena, mainly dual-tasking (doing two things simultaneously) and task switching (moving between tasks in sequence). Each has its own experimental paradigms, its own costs and its own neural signatures.

A 2018 integrative review in Psychological Bulletin makes exactly this point: research on dual-task interference and sequential task switching has largely proceeded separately, using different methods, even though both ask how humans handle multiple task demands¹. The review argues for treating them as related expressions of shared underlying cognitive mechanisms rather than isolated curiosities¹.

This piece breaks down what the studies actually show, study by study, so you can see which findings are robust, which are more limited, and what that means for how you work.

What is the "switch cost" and how reliable is it?

When researchers compare a trial where you repeat the same task against a trial where you switch to a different task, the switch trial is reliably slower and more error-prone. This performance penalty is the switch cost.

A 2010 review in Psychological Bulletin describes the switch cost as a general finding across the many procedural variations of the task-switching paradigm². The review notes that switching to a new task results in slower and more error-prone execution, and that part of this cost comes from resolving interference from the previous task, while another part comes from the mental work of reconfiguring your task set².

A second major review, also from 2010, covers the same ground from a different angle, detailing the interference phenomena that arise during task switching, including proactive interference from recently performed tasks and asymmetrical switch costs³. Together, these two reviews represent the strongest form of evidence available: large-scale syntheses of dozens of experiments, not single findings.

The practical reading is straightforward. Every time you move from one task to another, you pay a small tax in speed and accuracy. Individually the tax is tiny, often a few hundred milliseconds. Accumulated across a working day of frequent switching, it becomes a meaningful drag.

Does doing two things at once get worse if you also switch between tasks?

A 2025 randomised controlled trial asked a question that had not been properly tested before: does task switching add to the costs of dual-tasking, or is dual-tasking already hard enough on its own?⁴

Forty-five young adults aged 18 to 30 drove in a simulated rural landscape at a constant 70 km/h while performing additional typing and arguing tasks⁴. The additional tasks were presented either in separate, repetitive blocks or in intermixed, switching blocks. Driving performance was measured through average velocity and the standard deviation of lateral lane position, while the additional tasks were measured through reaction time⁴.

The result was more nuanced than the researchers expected. Performance was better in the repetitive condition than the switching condition only for one measure: the standard deviation of lateral lane position during the typing task⁴. On other measures, switching added little. The authors concluded that dual-tasking alone already involves processing demands that are not substantially increased by adding switching demands⁴.

This matters for how you think about your own work. It suggests the biggest cost comes from dividing attention at all, not from the specific way you divide it. A day spent doing one thing at a time, even if you switch between those things occasionally, may be less damaging than a day spent constantly trying to hold two things in mind at once.

What happens to memory when you dual-task as you age?

A 2025 randomised controlled trial examined cognitive-cognitive dual-tasking, meaning two mental tasks at once rather than a mental task plus a physical one⁵. Sixty-one healthy participants aged 50 to 77 completed a visual memory task as the primary task and a phonemic fluency task as the secondary task, either concurrently or one at a time⁵.

As expected, performance on both tasks declined under concurrent demands, and performance also declined with increasing age⁵. But the two factors did not interact: the dual-tasking cost was stable across the age range studied⁵. In other words, older participants were not disproportionately worse at dual-tasking than younger participants within this 50 to 77 age band, even though both age and dual-tasking independently reduced performance.

The study also found that participants with higher memory costs under dual-tasking performed worse on other measures of divided attention, such as the Trail Making Test, though no correlation was found with the MoCA cognitive screening score⁵. The authors suggest this pattern may point to a core ability for splitting attention across different types of information⁵.

For readers in midlife and beyond, the takeaway is not that multitasking becomes uniquely catastrophic, but that it carries a steady cost that sits on top of the normal cognitive changes of ageing. Protecting focus matters at every age.

Can practice erase the multitasking penalty?

A 2014 randomised controlled trial tested whether training could reduce dual-task costs in older and younger adults using a psychological refractory period paradigm⁶. Task 1 was a red/green signal discrimination, analogous to pedestrian signals, and Task 2 was a tone discrimination, analogous to traffic sounds⁶.

Participants completed a dual-task pretest, then six weeks of single-task practice, then a dual-task posttest⁶. Older adults showed larger overall dual-task costs than younger adults, but they were able to reduce those costs with practice just as much as the younger group⁶. However, even when single-task performance on Task 1 reached asymptotic levels, the dual-task costs did not disappear⁶.

The study also found something with real-world safety implications: older adults, but not younger adults, responded more rapidly to green "go" stimuli than to red "wait" stimuli, a bias the authors describe as potentially dangerous at crosswalks when combined with large dual-task interference⁶.

The practical lesson is that practice helps, but it does not solve the problem. You can get better at a specific task through repetition, yet the underlying cost of dividing attention remains. This is a strong argument for designing your environment so you do not need to multitask in the first place, rather than trying to train your way out of it.

Is "inhibition" one skill or several?

One reason multitasking research is messy is that the underlying cognitive abilities are not as unified as everyday language suggests. A 2004 study of 220 adults used confirmatory factor analysis to examine three inhibition-related functions⁷.

The results showed that prepotent response inhibition and resistance to distractor interference were closely related, but both were unrelated to resistance to proactive interference⁷. The authors argue that the term "inhibition" has been overextended and that researchers need to be more specific about which function they mean⁷.

This has a direct implication for how you interpret productivity advice. A technique that helps you resist a distracting notification may not help you stop an old task from intruding on a new one. They are related but separable abilities. When a tip promises to improve your "focus" in general, the research suggests you should ask: which kind of focus, and under what conditions?

Does the brain treat switching and maintaining differently?

A 2003 fMRI study decomposed brain activity during task switching into sustained and transient components⁸. Contrasting task-switching blocks against single-task blocks revealed sustained activation in the right anterior prefrontal cortex, while contrasting switch trials against repeat and single-task trials revealed activation in the left lateral prefrontal cortex and left superior parietal cortex⁸.

Crucially, the right anterior prefrontal cortex activity selectively covaried with the magnitude of the mixing cost (the cost of being in a block where switching is possible), while the left superior parietal activity selectively covaried with the switching cost itself (the cost of an individual switch)⁸. The authors describe this as a functional double dissociation⁸.

What this means in plain terms: your brain has one system for holding a task set in place over time, and a different system for the act of switching between tasks. Multitasking is not a single mental muscle. It is a coordination problem between multiple control systems, which helps explain why the costs are so hard to train away entirely.

Does task switching develop differently from other control skills?

A 2006 clinical trial tested 325 participants aged 4 to 13, plus young adults, on a computerised battery designed to measure working memory, inhibition and cognitive flexibility⁹. The design manipulated memory and inhibition both independently and together, in single-task blocks and during task switching⁹.

The findings showed a striking developmental pattern. Even the youngest children could hold information in mind and inhibit a dominant response when the rules stayed constant⁹. But cognitive flexibility, the ability to switch between rules, showed a much longer developmental progression, with 13-year-olds still not at adult levels⁹.

The study also found that costs associated with inhibition were greater in young children, while costs associated with increasing memory demands were greater in adults⁹. Adults slowed down on difficult trials to preserve accuracy, whereas the youngest children were more impulsive, keeping their response times constant but paying an accuracy cost⁹.

This matters if you manage or parent young people, or if you simply want to understand your own attention. The ability to switch between tasks is one of the slowest-developing cognitive skills, which means expecting a teenager to handle constant task switching the way an adult would is unrealistic.

What about people who switch between languages daily?

Bilingual speakers offer a natural experiment in task switching, since they constantly manage two language systems. A 2013 observational paper proposed the adaptive control hypothesis, which argues that language control processes themselves adapt to the recurrent demands placed on them by the interactional context¹⁰.

The authors distinguish eight control processes, including goal maintenance, conflict monitoring, interference suppression, selective response inhibition, task disengagement and task engagement¹⁰. They predict that different interactional contexts, such as a single-language context versus a dual-language context versus dense code-switching, place different demands on these processes and lead to adaptive changes in the associated neural circuits¹⁰.

This is a hypothesis paper rather than an experimental finding, so it should be read as a framework for future research rather than settled evidence. But it offers a useful perspective: the demands you place on your attention regularly may shape the control systems that serve it. If you spend your days in a dense code-switching environment, you are exercising a different set of control processes than someone who works in long, uninterrupted blocks.

Studies at a glance

What this means for how you work

The research does not support the idea that some people are simply "good at multitasking" in a way that lets them escape the costs. The costs are measurable, robust across study designs, and present even after practice. What varies is the size of the cost, not its existence.

The single most useful shift is to reduce the number of times you divide your attention at all. The 2025 driving study suggests that the jump from single-tasking to dual-tasking is the expensive one, and that adding switching on top of dual-tasking adds comparatively little⁴. So the priority is not to become a better switcher, but to spend more time not dividing your attention in the first place.

A sensible starting point is to pick one task, set a timer for 25 minutes, and work with zero context switching. This is the Single-Task Sprint from the Unleashed Moves Bank. Pair it with keeping one task visible on screen at a time, using full-screen mode or a single window, a technique from the Unleashed Environment Bank. If your attention feels scattered before you begin, the Visual Focus Training move from the Unleashed Moves Bank suggests staring at a single point for 60 to 90 seconds to settle your attention before a work session.

These are practical techniques, not research findings, but they align with what the studies imply: the fewer demands you place on your attention at once, the better you perform.

Bottom line

The evidence that single-tasking beats multitasking is strong. Multiple large-scale reviews confirm that task switching produces reliable speed and accuracy costs^2,3, and randomised controlled trials show that dual-tasking degrades performance on both tasks^4,5. The evidence that practice can eliminate these costs is weak: training reduces but does not remove them⁶. The idea that multitasking ability is a single trainable skill is not supported; the cognitive and neural evidence points to multiple, partially separable control systems^7,8.

For most knowledge work, the practical implication is clear. Structure your day around protected blocks of single-task work. Treat multitasking not as a skill to master but as a tax to minimise. If you want to go deeper on the practical side, see our guide on why you cannot focus for more than a few minutes and our piece on how to focus when everything feels urgent.