Studying creative work in the 1960s, Mihaly Csikszentmihalyi noticed that painters kept going single-mindedly while a picture was going well, ignoring hunger and tiredness, then lost interest in the finished canvas almost at once. Whatever was rewarding was happening during the work. He went on to interview chess players, rock climbers and dancers, and to study surgery.¹

A flow state is what he found they were describing: complete absorption in an activity that feels worth doing for its own sake. You cannot switch it on. What the research offers instead are conditions that make absorption likely, and a fairly sober account of what it is worth. A 2023 meta-analysis of studies of flow at work found that employees who report more flow also score higher on job performance, but most of those studies measured everything once, so they cannot show which came first.²

What people mean by a flow state

Flow is complete absorption in the present moment. Jeanne Nakamura and Mihaly Csikszentmihalyi describe the state as intense concentration on what you are doing, action and awareness merging, a loss of awareness of yourself as a social actor, a sense that you can handle whatever comes next, time passing differently, and an activity that feels rewarding in itself.¹

The reason those go together is attention. In the model Csikszentmihalyi built alongside the flow concept, information reaches awareness only through the selective investment of attention, and attention is limited. Put all of it into a set of demands you can just about meet and there is none spare for the running commentary most of us carry around: how am I doing, how much is left, what will they think.¹

Information appears in consciousness through the selective investment of attention.

You have probably had a version of it. A translator working paragraph by paragraph through a document that visibly shortens looks up expecting mid-morning and finds the afternoon gone. Nothing about that is mystical. She had one thing to do, she could see it being done, and it was hard enough to need her.

What this means for you: those six features describe an experience, not a set of instructions. None is an action you can take.

Myth

With the right technique you can drop into a flow state whenever you need one.

Fact

Flow theory describes conditions that make deep absorption likelier. The absorption itself is the readout, not the move.

If you want the broader habits that protect concentration in an ordinary week, start with how to focus on what matters. Flow is the narrower question of what happens when concentration goes all the way.

Two things you arrange, six things you notice

The model splits in half, and only one half is under your control. Nakamura and Csikszentmihalyi list two conditions of flow: perceived challenges that stretch your existing skills without overmatching them, and clear near-term goals with immediate feedback about the progress you are making. Everything else on the list is what people report once they are in it.¹

The famous “nine dimensions” are a questionnaire’s way of counting. Susan Jackson and Herbert Marsh, building a flow measure for sport, wrote nine scales of four items each to represent the dimensions Csikszentmihalyi had described, and administered the 36-item result to 394 athletes.³ In a 2022 opinion article critiquing the scales that grew out of that work, Joy Lee-Shi and Robert Ley of Simon Fraser University describe the nine as falling into two categories: three conditions, which are challenge-skill balance, clear goals and unambiguous feedback, and six subjective experiences.⁴ Same split, with the goals-and-feedback line counted as two.

Checking whether you feel absorbed is itself the running commentary that absorption rules out.

Reading the nine as one flat list is what causes the trouble, because a flat list hides the difference between the levers and the readings. A rating scale has to score everything on the same axis; your morning does not. You can decide to give the next hour to one document. You cannot decide to lose track of time while you do it.

So work on the first two rows and leave the rest alone. Checking whether you feel absorbed is itself the running commentary that absorption rules out.

Where the famous flow diagram came from

The diagram with flow running up a diagonal channel between anxiety and boredom is a model, not a measurement. Csikszentmihalyi drew the first version from his interviews in the 1970s. When researchers in Italy paged people at random through ordinary days and asked what they were doing and how it felt, the results did not match it, and the map was redrawn.¹

The problem was balance on its own. Nakamura and Csikszentmihalyi report that simply balancing challenges and skills did not produce the best experience: an evening of television matches low demands to low skill, and it is balanced, and nobody would call it flow. The Milan group redefined flow as the balance of challenges and skills when both are above average levels for that person, which opened up a fourth region of the map, apathy, where low challenge meets low skill.¹

Skill below your averageSkill above your averageChallenge above your averageChallenge below your average

Challenge above your average, Skill below your average: Anxietythe work is ahead of you

Challenge above your average, Skill above your average: Flowstretched, and able to meet it

Challenge below your average, Skill below your average: Apathylittle is asked, little is brought

Challenge below your average, Skill above your average: Relaxationeasy, pleasant, not absorbing

So “matched difficulty” is a weaker test than it sounds. Filing expenses is well matched to most people’s skill and rarely absorbs anyone. The question is whether the task asks more of you than an ordinary hour does, and whether you have the skill to meet what it asks. That is also why absorbing work and work that produces something original tend to look alike: both sit near the top of what you can currently do.

Further reading

• 

  Flow: The Psychology of Optimal Experience

  by Mihaly Csikszentmihalyi

  Csikszentmihalyi's own book-length account of flow, from the researcher whose work this article draws on.

As an Amazon Associate WiserHours earns from qualifying purchases.

How to get into a flow state

You arrange the conditions and let the state do what it does. Nakamura and Csikszentmihalyi name two: a challenge that stretches your existing skills, and clear near-term goals with immediate feedback on progress.¹ Split the second in two, as the nine-dimension questionnaires do, and you have three things to set up before a long stretch of work.⁴ Only the first has an experiment behind it in the studies cited here.

1. Matched difficulty: more than an ordinary hour asks, within the skill you have
2. One goal: name what finished looks like for the next stretch of work
3. Visible feedback: something that changes while you work, so progress shows

1. Pitch the work above what an ordinary hour asks

This is the condition with experimental support behind it. Johannes Keller and Herbert Bless report two studies that depart from correlational research and use a newly developed experimental paradigm to document, in their words, the causal impact of the compatibility of a person’s skills and a task’s demands on the emergence of flow. We read their abstract only, so we take nothing from the paper beyond that.⁵

The practical move is to change the task rather than yourself. A support agent whose queue has gone routine sets a harder mark: clear the three oldest tickets before lunch, writing each reply so the next person can reuse it. Someone avoiding a long report does the opposite and cuts it down to the one section they already know how to start. Either way the hour asks a little more than usual without asking more than you have. When a mismatch shows up as postponing the work instead of starting it, the more useful reading is what makes us put things off.

2. Give the stretch of work one goal you could say out loud

Office work rarely arrives with a finish line. “Work on the strategy document” is a direction, not a goal: it is still true at six o’clock, however the afternoon went. “Two pages of the options section, rough is fine” tells you when to stop. Karen Bartholomeyczik and colleagues, reviewing the work on flow at work, describe a study in which workers prompted on five consecutive mornings to write down three goals for the day reported more flow, and note that its authors could find only one other empirical test of a flow intervention at work.⁶

3. Pick something that visibly changes while you work

Feedback is the condition office work drops most often. A climber gets it from the next hold; a document says nothing back until somebody reviews it. So borrow a substitute: a word count, sections ticked off a list, a test suite going green, a queue with a number on it. It does not have to measure quality. It has to move while you work, so progress is something you see rather than something you stop and estimate.

Then protect all three, because any of them can be met and still lose to a tap on the shoulder. Bartholomeyczik and colleagues include reducing interruptions among the practical ways to support flow at work, and suggest quiet workspaces that prevent them.⁶ What a switch leaves behind afterwards, and how to soften it, is the subject of attention residue.

Run that setup a few times before deciding it does not suit you. Flow at work varies a lot between people and from day to day, which is part of why Bartholomeyczik and colleagues argue that interventions have to adapt to the person rather than apply one recipe.⁶

WiserHours

The setup, in five lines

• One task chosen, harder than an ordinary hour asks
• A finish line I could tell someone in a sentence
• Something that visibly changes as I work
• Messages and calls off for the block
• Judged afterwards by the work, not by how absorbed I felt

Find this checklist online:
https://wiserhours.com/focus/flow-state/#checklist-flow-state-setup

Does a flow state actually make you work better?

Probably a little, and the evidence is weaker than the claims made for it. Employees who reported more flow also scored higher on job performance, work engagement and job satisfaction, in a meta-analysis of studies of flow at work published in 2023 by Wei Liu and colleagues. Those authors state plainly that their analysis cannot establish cause.²

The study

Limited evidence

Flow went with better job performance, and hardly any of it was experimental

Pooling studies of flow at work, Wei Liu and colleagues report a positive relationship with job performance, and stronger ones with work engagement and job satisfaction. They report that the analysis cannot address causal effects, because most of the studies it pools measured flow and its outcomes at the same moment.²

A correlation that size is the kind you would notice across a whole workforce and easily miss in one person. It is also open at both ends: flow may help people work well, or working well may be what produces flow. A week that happens to be going smoothly for reasons of its own would produce both ratings at once.

One study did separate the two measurements in time. Stefan Engeser and Falko Rheinberg asked 246 German psychology students, most of them women in their early twenties, to rate their flow during a short study session a week before a compulsory statistics exam. The flow they reported then explained about 4 percent of the differences in their final marks, over and above age, school math grades and prior knowledge.⁷

The size is the lesson. In the same analysis, age and school math grades each accounted for more of the difference in marks than flow did, and in a round of Pac-Man run by the same researchers the relationship was smaller and not clear.⁷ Treat absorption as one contributor to how a piece of work goes, sitting well behind skill: worth arranging before a hard task, not worth rebuilding your week around.

How much to trust a claim about flow

Cautiously, and mainly because of how flow is measured. In a 2020 review, Sami Abuhamdeh searched the previous five years of the psychological literature and found 42 studies that measured flow in 24 different ways, often differing so much that the word changed meaning between papers.⁸ When the construct moves like that, results across studies cannot be compared.

So when one article reports that flow lifts output and another reports that it changes nothing, the useful question is what each one measured, and the answer is usually a different questionnaire asked at a different moment. Cause is the second open question, and the Liu meta-analysis names it: its own authors call for experimental, intervention and longitudinal designs to settle what flow does.² Whether anyone can produce flow on purpose is the third. Bartholomeyczik and colleagues found so few tested workplace interventions that their paper is a framework for building them rather than a verdict on any, and they add that being in flow all the time cannot and should not be the goal.⁶

None of that makes the idea useless. It makes the promise smaller and more honest: the conditions are real, they are arrangeable, and the state arrives sometimes. Anyone selling a reliable route into it is ahead of the evidence.

The bottom line

You cannot enter a flow state on command, and the research saying flow makes you better at your job is mostly people rating themselves once, in one sitting. What survives is worth doing anyway. Choose one task that asks more of you than an ordinary hour, give it a finish line you could say out loud and a result you can watch change, and take the interruptions away. Absorption is what happens when those hold, not something you can summon first.