Zoom Fatigue: The Famous Explanation Was Never a Study

You have almost certainly encountered the four causes of Zoom fatigue: too much close-up eye contact, the cognitive effort of sending and reading nonverbal signals through a screen, the strain of seeing yourself all day, and being pinned in one spot.

They come from a 2021 paper by Jeremy Bailenson at Stanford. It is one of the most cited things ever written about video calls, and it contains a sentence that almost nobody who quotes it repeats:

“All are based on academic research, but readers should consider these claims to be arguments, not yet scientific findings.”

Bailenson said, in the paper itself, that his four causes had “yet to be directly tested in the context of Zoom, and require future experimentation to confirm.”

That was five years ago. The experiments have since been run. This article is about what they found - because the answer is more interesting than the argument, and it is not the same as the argument.

First: the tiredness is real, and it is measurable

Before anything else, one thing is now settled that was not settled in 2021. Video calls really do produce a physiological state that looks like fatigue. This is not a self-report artefact or a pandemic mood.

Riedl, Kostoglou, Wriessnegger and Müller-Putz (2023) ran the cleanest possible comparison in a laboratory. Thirty-five students attended the same 50-minute university lecture twice - once by videoconference, once face to face, counterbalanced - with EEG and ECG recording throughout.

Theta power at Fz was larger during videoconferencing and rose steadily across the lecture (slope = 2.7240 × 10-6, p = 0.0145). Alpha power was more pronounced in the videoconferencing condition. The theta/beta ratio was significantly greater during videoconferencing (p < 0.05).

Their conclusion: 50 minutes of videoconferencing, compared with face to face, “results in changes in the human nervous system which… can undoubtedly be interpreted as fatigue.”

Same lecture. Same lecturer. Same students. Different medium, different brain state.

The one cause that has been nailed down

Of Bailenson’s four, exactly one has been isolated and confirmed by two independent methods. It is the one about seeing yourself.

In a classroom, across a semester. Basch, Albus and Seufert (2025) ran four weekly intervention comparisons in a real university online course. Switching self-view off produced the largest effects they measured:

In an EEG lab. Xu, Whelan, O’Brien and O’Hora (2024) put 32 people through live video meetings with self-view toggled on and off, 20 minutes per condition. Their stated aim was blunt: previous research was “informed primarily by self-reported data,” and they wanted “neurophysiological evidence.” They found “significantly greater alpha activity when self-view was on than when it was off.”

Two different methods, two different countries, one finding. Turn off your own video preview. It is the single best-supported piece of advice in this entire literature, and it takes one click.

Notice also what the same table rules out. Grid view versus speaker view - a thing people argue about - did not reach significance. A virtual background did not significantly change fatigue, though it did raise extraneous cognitive load.

The part of the story that did not survive

The most-repeated corollary of the Zoom fatigue thesis is that women suffer more, and that self-view is why.

It came from a large survey by Fauville, Luo, Queiroz, Bailenson and Hancock (2021): 10,591 respondents, in which “women (M = 3.13, SD = .78) reported a significantly higher level of Zoom fatigue than men (M = 2.75, SD = .81) (t(5506) = -21.9, p < .001, d = .48)” - reported in the paper as 13.8% higher.

Two things have happened to that finding since.

Its own replication shrank it. The same paper’s replication sample of 778 found the difference held but at d = .16 - about a third of the original effect.

The EEG test did not find it. Xu et al., having confirmed the self-view effect neurophysiologically, reported that alpha activity “was not significantly different for men or women.” The mechanism replicated. The gendered version of it did not.

That is not a reason to dismiss the survey - self-reported exhaustion is a real thing to measure, and it may differ for reasons the EEG cannot see. It is a reason to stop repeating “women experience more Zoom fatigue because of self-view” as an established finding with a mechanism attached. The mechanism half has now been tested and did not appear.

What actually predicts fatigue may not be about faces at all

The largest synthesis to date is Beyea and colleagues (2024), a meta-analysis of 56 antecedents across 38 quantitative studies. Its headline is not a nonverbal mechanism:

“We observed the largest effects on VF within the psychological factors category (r = 0.24, k = 28), with ‘feeling trapped’ (r = 0.33, k = 5) being the largest predictor among uncategorized variables.”

The strongest single predictor of videoconference fatigue in the pooled literature is feeling trapped. Not gaze distance, not self-view, not grid size - the sense of having no way out of the call.

Two honest cautions on that. The effect sizes are modest: r = 0.24 for the strongest category is a real but not dominant relationship. And “feeling trapped” rests on only five studies.

But it points somewhere Bailenson’s framework does not. Three of his four causes are about the image. This one is about autonomy - whether you chose to be there, whether you can leave, whether the meeting could have been an email. If it holds, the most effective intervention is not a camera setting. It is a shorter meeting you are allowed to decline.

What “Zoom fatigue” is, and what the measure actually measures

The standard instrument is the Zoom Exhaustion & Fatigue Scale, built by Fauville and colleagues across five studies with, by their own summary, “over 700 participants.” It measures five dimensions - general, social, emotional, visual and motivational fatigue - with all subscale alphas above .8.

It is a good scale. It is worth being precise about what it does: it measures self-reported fatigue. It does not test causes. A great deal of writing on this subject cites the scale paper as though it explained the phenomenon. It quantified it.

Riedl’s separate 2022 conceptual paper, built on a systematic review of 45 articles published before May 2021, found that only 12 of them offered an explicit definition of Zoom fatigue at all. That is the state the field started from: a widely felt experience with no agreed definition, one influential argument about its causes, and very little testing.

The wider question of what a screen does to sustained attention sits in the screens, attention and focus guide; for the cost of moving between tasks rather than sitting in one, see lost work time and task switching.

What this does not say

It is not a medical claim. Nothing here is a diagnosis, and “fatigue” in this literature means measured EEG and ECG changes and self-reported tiredness after a call - not a clinical condition. If you are persistently exhausted, that is a question for a doctor, not for an article about video calls.

It does not say video calls are worse than meetings. The comparison in the strongest experiment was a lecture delivered two ways. Nobody has shown that a video call is more tiring than the in-person meeting plus the commute to it.

It does not say the other three causes are false. Close-up gaze, nonverbal load and physical constraint remain plausible and largely untested. “Not yet demonstrated” is not “disproved” - and Bailenson said as much at the time.

The evidence is young and the samples are small. Riedl’s experiment had 35 participants; Xu’s had 32; Basch’s per-comparison numbers ranged from 12 to 28 in a field study with voluntary attendance and real attrition. These are the best experiments that exist, and they are small.

What to do with this

Turn off self-view. It is the only one of the famous four with two independent confirmations behind it, and the effect sizes are the largest in the literature.

Then ask whether you needed to be in the call. The pooled evidence points at “feeling trapped” more strongly than at anything on screen, which suggests the problem is often the meeting rather than the medium.

And when you next see the four causes listed as fact - in a memo, a wellness email, a management article - remember that the man who proposed them wrote, in the same paper, that they were arguments and not yet findings. Five years later, one of them is a finding. That is a real result, and it is a smaller and more useful claim than the one that got famous.