Should You Smile in a Headshot? The Mechanics of Getting a Real One

Should you smile in a headshot? Usually yes — but the smile has to reach your eyes. A mouth-only smile reads as strained at the small sizes profile photos actually render at, while a genuine one signals warmth and competence at once. The harder question isn't whether to smile. It's how to produce one on cue.

Almost every article on this question answers it with taste: smile, it's friendlier, or don't, it undermines authority. Taste is a bad guide here, because the difference between a headshot that works and one that doesn't is rarely the decision to smile. It's the execution — which muscles moved, how long the expression had been held, and whether any of it is still legible once the image is scaled down to the size of a fingernail.

The two smiles your face can make

Anatomically there are two of them. One is driven by the zygomaticus major, the muscle that pulls the corners of the mouth up and out. The other adds the orbicularis oculi, the ring of muscle around the eye socket that narrows the eye opening and creases its outer corner. Only the second one reads as felt.

That second version is the Duchenne smile, named after the 19th-century French neurologist Guillaume Duchenne, who noticed that the mouth could be commanded but the eye muscle seemed to answer only to real feeling. The perceptual half of his claim has held up well. One elegant test photographed people smiling before and after botulinum toxin treatment for crow's feet — same person, same instructions, with the eye muscle chemically quieted in the second set. Raters judged the pre-treatment smiles, the ones with orbicularis oculi still active, as more felt, more spontaneous, more intense and happier. The mouth-only version — eyes wide and static — is what people mean when they say a photo looks forced, even when they can't say why.

Here's where the popular version of this story overreaches, and it's worth being honest about. The eye crinkle is not a lie detector. Later research has walked back the idea that orbicularis oculi is beyond voluntary control: in one imitation task run by Gunnery, Hall and Ruben, around 71% of participants managed to reproduce a Duchenne smile on request. Actors do it professionally. So the eye component isn't proof of sincerity. What it is, reliably, is the difference between an expression that looks engaged and one that looks applied. For a photograph, that's the only thing that matters — and it also means the eye engagement is something you can learn, rather than something you either feel or don't.

There's a second, less discussed reason the eyes carry the load: a mouth can be held. An eye expression can't. Hold a smile for eight seconds and the mouth stays put while the eye region quietly relaxes back to neutral. That decay is what produces the classic bad headshot — a wide, symmetrical, perfectly pleasant mouth sitting under two flat, slightly anxious eyes. Nothing is wrong with the smile. It has simply outlived the feeling that started it.

What the research actually says about smiling and being taken seriously

The popular advice splits into two camps: smile and look friendly, or stay neutral and look serious. The published research is messier than either camp. It suggests smiling helps more often than it hurts, that intensity matters more than presence, and that the effects are real but modest rather than career-defining.

Three findings are worth knowing, because together they explain why the answer isn't a simple yes.

1. 1Smiling generally reads as more competent, not less. A 2020 study in the Australian Journal of Psychology by Filkuková and Jørgensen asked 238 employees of IT companies to rate photographed candidates for a software developer role. Each rater saw one of three expressions: smiling, neutral, or a hand-on-chin "thinking" pose. Smiling faces were rated the most competent, neutral came second, and the thinking pose came last.
2. 2But bigger is not better. A 2017 Journal of Consumer Research paper by Wang, Mao, Li and Liu tested smile intensity directly and found a trade-off: a broad smile made a person seem warmer than a slight smile, but less competent. The warmth gain dominated when the stakes for the viewer were low; the competence cost bit harder when the purchase was risky or the viewer was focused on avoiding a bad outcome. That is the single most useful finding for headshots, because it reframes the question from smile/no-smile to how much.
3. 3The real-world effect sizes are small. A large field study of Airbnb listings by Zhang, Friedman, Srinivasan, Dhar and Zhang, published in the Journal of Consumer Research, used a machine classifier to detect smiles in host profile photos and found them associated with roughly a 3.5% lift in demand on average — meaningful at scale, but not the transformation some headshot marketing implies. The effect was concentrated among male hosts, where it exceeded 8%, and was not statistically significant for female hosts. It was also larger where a guest faced more uncertainty, which is a useful hint: the photo does the most work when the viewer knows the least about you.

Commercial photo-rating platforms report the same direction from far larger but far less controlled datasets. Photofeeler's aggregate ratings show open-mouth smiles outperforming closed-mouth smiles, and both outperforming non-smiling photos, on business-oriented traits — with likability moving most and competence and influence moving less. Treat those as directional vendor data rather than science. They're consistent with the peer-reviewed work, which is the most you can say for them.

One uncomfortable detail from the same 2020 study deserves mention: women in the sample were rated as less competent than men overall, an effect the authors found was driven by raters without prior hiring experience. No expression choice fixes that, and it's a reminder that the photo is one input among many rather than a lever with unlimited travel. We looked at the broader version of that question — what a photo does and doesn't control in hiring — in our piece on age bias, LinkedIn photos and what you actually control.

Teeth or no teeth

Teeth or no teeth is the part people agonize over, and it's the part where personal comfort should carry the most weight. Both work. A closed-mouth smile that engages the eyes beats a wide toothy one that doesn't, and a strained closed mouth loses to an easy open one. Comfort is the tiebreaker because discomfort is visible.

When an open smile reads better

• Client-facing and volume-contact roles — sales, recruiting, real estate, hospitality, coaching. The first job of the photo is to lower the cost of contacting you.
• When your resting face reads as severe. Plenty of people have a neutral expression that photographs as displeased. If yours does, a closed mouth compounds the problem rather than fixing it.
• Speaking and press use. Conference sites and press pages sit your photo next to a dozen others; an open expression separates you from a wall of composed neutrals.
• When the smile is genuinely yours. Some people simply don't have a natural closed-mouth smile — the mouth-corner-only version looks like a smirk on them. Don't manufacture one.

When a closed-mouth smile reads better

• High-stakes advisory contexts — litigation, wealth management, surgery, board-level consulting — where the competence trade-off documented in the smile-intensity research is worth paying attention to.
• When you're self-conscious about your teeth. Self-consciousness leaks. A relaxed closed smile always outperforms a guarded open one.
• When you photograph with a very wide smile. Some faces go from pleasant to laughing in a very short range of mouth opening. If yours does, the closed version gives you more control.
• For a second frame in your set. Even if the open smile is your primary, a closed-mouth alternative is the frame you'll want for a serious context later.

The genuinely useful middle option gets overlooked: the parted-lip smile, where the lips separate slightly and the top teeth show a little without the full grin. It sits right in the range the intensity research points to — warm enough to register, restrained enough not to cost you authority — and it survives downscaling better than a closed mouth, because a thin line of light between the lips gives the eye something to read.

How to produce a real expression on cue

Here's the practical problem. Knowing you want a Duchenne smile doesn't give you one. Expression isn't a position you hold; it's a state you pass through. Every reliable technique works the same way — it hands you something real to react to a fraction of a second before the shutter, rather than asking you to perform an outcome.

Reset between every frame

This is the single highest-return habit and almost nobody does it. Don't hold an expression across a burst. Drop the face completely to neutral, exhale, then let the expression rebuild for one or two frames. Then drop it again. A photographer shooting continuously for thirty seconds is mostly documenting expression decay. Three cycles of reset-and-rebuild will produce more usable frames than three minutes of holding.

The countdown that lands on the wrong number

Photographers count to three and shoot on four. The reason is that the expression you construct on "three" is the deliberate one; what follows half a second later — the small release when you think the moment has passed — is the real one. If you're shooting yourself on a self-timer, set it to ten seconds, get your expression ready at six, and let the shutter catch you at the point where you've stopped trying.

The fake laugh

The oldest trick in portrait work: the subject is asked to produce a deliberately terrible fake laugh on a count of three. Nobody photographs the fake laugh. They photograph the two seconds after it, when the absurdity of having just done it produces an actual one. It works because it converts an instruction you can't follow ("look happy") into one you can ("make a silly noise").

Think of a specific person, not a happy thought

"Think of something that makes you happy" is too abstract to move a muscle. Specificity is what works. Picture one particular person walking into the room right now — a friend you haven't seen in a year, a small child you know, someone whose arrival would genuinely lift your mood. The face responds to a concrete image in a way it never responds to a category.

Talk through the shutter

Speaking mobilizes the whole lower face and prevents the mouth from setting. Photographers keep subjects mid-sentence for exactly this reason. Alone with a phone, say something out loud — a sentence, a name, anything — and shoot while you're talking. Roughly one frame in five will be an unusable mid-consonant mouth shape. The other four will be looser than anything you can pose.

Engage the lower lid deliberately

If the eyes are the part that goes flat, work them directly. Portrait photographer Peter Hurley popularized the squinch: raise the lower eyelid slightly toward the pupil while letting the upper lid come down only a hair. It's not a squint — a squint closes both lids and kills the eye. Done correctly it narrows the eye opening mostly from below, which produces the settled look of someone comfortable rather than the wide, alert eye of someone bracing for a flash. Practice it in a mirror before you practice it on camera; the sensation is unfamiliar and the first few attempts usually overshoot.

What to stop doing

• Saying "cheese." The long ee stretches the mouth laterally into a shape no genuine smile makes, and it does nothing to the eyes. Any word ending in an open vowel sound is better.
• Watching yourself on the screen. Self-monitoring flattens expression instantly. Flip the phone to the rear camera or look away from the preview.
• Shooting only once. Expression is a numbers game even for professionals. Forty frames across three or four resets is a normal working set, not excess.

If the whole exercise makes you tense — which is the ordinary reaction, not a personal failing — the body mechanics matter as much as the face. Our guide to posing for a headshot when you're not photogenic covers the shoulder, chin and weight adjustments that make a relaxed expression physically easier to hold.

What happens at thumbnail size

Almost nobody sees your headshot at full size. LinkedIn accepts profile photos from 400 × 400 pixels up to 7680 × 4320, at a maximum of 8 MB — but 400 is the floor, not the target, and what matters more is the other end of the scale. Your profile page shows the photo a few hundred pixels wide; in a feed, a comment thread or a search result it collapses to roughly fifty pixels. Design for the small end.

At a few hundred pixels most expressions survive intact. At fifty, the image has been reduced to something closer to a symbol than a photograph, and only a few things still carry:

• The dark shape of the eyes. This is the most durable feature in the frame. A narrowed, engaged eye keeps a distinct shape at tiny sizes; a wide, flat eye becomes a featureless dot.
• The overall curve of the mouth. The direction survives — up or flat. The detail does not. Teeth stop being teeth and become a pale band.
• Head-to-background contrast. If your hair and background sit at the same value, the silhouette dissolves and the whole thumbnail turns to mush regardless of how good the expression was.
• Nothing else. Not skin texture, not a subtle asymmetry, not the careful thing you did with one eyebrow.

There's a practical consequence people miss. Because teeth collapse into an undifferentiated bright band at thumbnail size, a very wide open smile can end up reading as a white blob in the lower third of a fifty-pixel circle — the opposite of the polished impression it made at full resolution. The parted-lip and moderate open smiles hold up considerably better. So does a closed-mouth smile with genuine eye engagement, which is the version most people underrate.

Two secondary effects are worth knowing about. The first is the circular crop: every major platform masks a square upload into a circle, so the corners are gone and anything you were relying on at the edge of the frame — a shoulder line, a hand, a sliver of background — may not survive. The second is compression. Platforms re-encode what you upload, and heavy retouching or a busy background gives the encoder more to lose. A clean, well-separated portrait degrades more gracefully than a detailed one.

Test this before you commit. Export your candidate at around 50 × 50 pixels, or just shrink the browser window until the photo is thumbnail-sized, and look at it from across the room. Whatever's still legible is what your network actually sees. If you want the full set of platform dimensions, crops and safe areas, we keep them current in our LinkedIn photo size and specs reference.

How the answer shifts by field

Field norms are real, but narrower than people assume. Almost no profession requires a stone face; what changes is the ceiling on intensity and how much of the frame the smile is allowed to occupy. Use the ranges below as calibration rather than as rules, and check them against the actual photos on your own firm's website.

Before the list, one boundary worth stating plainly, because these two questions get tangled constantly. This article is about producing an expression — the anatomy, the technique, the frame you can actually get. The strategic question of which signal to send in the first place — how far to lean toward warmth versus authority, and what that trade costs you in a given role — is the subject of our separate guide to the approachable versus authoritative headshot. Read that one to decide the target; read this one to hit it.

Law, finance and professional services

The most conservative end, and the field where the competence-versus-warmth trade-off is most live, because clients are making high-stakes decisions. The convention is a slight to moderate smile — closed or barely parted — with clearly engaged eyes. Deadpan neutral is not the norm on most firm bio pages any more; scroll a partner directory and you'll find far more soft smiles than blank stares. What you won't find is a laugh.

Sales, business development and recruiting

The one area where erring toward a bigger smile is straightforwardly correct. Your photo's function is to reduce the friction of a cold reply. Open smile, teeth, eyes fully engaged. The intensity research's competence penalty is a poor trade here, because warmth is what the role is actually being evaluated on.

Healthcare and clinical work

The hardest calibration in the list, because patients need reassurance and confidence in roughly equal measure. A 2025 study in Frontiers in Medicine that coded medical students' faces during simulated consultations found smiling predicted how comfortable standardized patients felt — and, notably, that this held whether the smile was genuine or merely social. That's an argument for warmth, but it describes a live consultation rather than a still photo, where a smile that reads as breezy can undercut clinical seriousness. The working answer is a warm, composed, moderate smile — the expression you'd want walking into a consultation room, not the one from a staff party.

Creative, media and acting

The widest permitted range, and the field where a genuinely neutral expression is most defensible — because in creative work the photo is a piece of positioning, not just an identifier. Acting is a special case with its own rules entirely: casting is reading you for castability, not likability, and the conventions diverge sharply from corporate work. We set out the differences in acting headshot versus corporate headshot.

Academia and research

Institutional pages skew neutral-to-slight, and the local convention on your department's faculty page is the only benchmark that matters. Worth noting: the 2020 competence study found the hand-on-chin "thinking" pose rated lowest of the three expressions tested. The pose that most reads as intellectual to the person striking it reads as staged to everyone else.

Real estate and local services

Open smile, and don't overthink it. These photos appear on yard signs, portal listings and business cards at wildly different sizes, and warmth plus legibility beats subtlety every time. Subtle expressions are the first casualty of a photo printed at two inches and viewed from a moving car.

Smiling in an AI headshot: what carries over and what doesn't

AI headshot generators learn your face from the selfies you upload, which makes expression one of the few things you still influence directly at the input stage. That's a real advantage and a real limitation, and it's worth being precise about which is which rather than pretending the technology has no seams.

What works in your favor: because a generator produces dozens of variations, you get a range of expressions to choose from in one pass rather than one photographer's best guess at your mood on a Tuesday morning. Selection pressure is genuinely useful — with fifty frames, the odds that several land in the warm-but-composed band are good. It also removes the performance anxiety that ruins a lot of studio sessions, since you're smiling at your own phone rather than at a stranger with a light stand.

What doesn't carry over, and where a good photographer still wins: a photographer is running a live feedback loop. They can see the eye region going flat, tell you a bad joke, and catch the recovery. A model can't do that. It generates from what your input photos taught it, which means if every selfie you upload has the same mouth-only smile, that's the smile you'll get back, fifty times over. The generator is not going to invent an engaged eye you never gave it. If your expression range matters more than turnaround time or cost, that is a genuine argument for booking a session with a person.

Two practical consequences follow. First, vary the expressions in your inputs deliberately — some neutral, some closed-mouth, some genuinely amused, ideally shot on different days. Our guide to taking input selfies that actually work covers the full set of requirements. Second, know that the eye region is exactly where generative models are weakest: mismatched catchlights, a smile that stops at the cheekbones, or teeth rendered with slightly too much regularity are the common tells. We cataloged them in ten signs an AI headshot looks AI-generated. Check the eyes at full resolution before you publish anything.

How to choose the frame you actually ship

You'll end up with more usable frames than you need, and choosing badly undoes the work. The common failure is picking the photo you like as a picture of yourself rather than the one that does the job at thumbnail size, for the audience it's aimed at. Those are rarely the same image, and the gap is where most people go wrong.

1. 1Cut anything where the eyes are flat. Do this first, before considering anything else. It usually removes half the set in one pass.
2. 2Shrink the survivors. View the remainder at thumbnail size. Anything that turns ambiguous or blank is out, regardless of how good it looked large.
3. 3Check the smile against the context. A litigation bio and a sales profile don't want the same frame from the same set. Pick per placement, not once for everything.
4. 4Get a second opinion from someone who doesn't know you well. People close to you correct for what you're "really" like. Strangers see only the photo, which is what your audience is doing too.
5. 5Keep two. One warmer, one more composed. You'll need both within a year, and going back for a second shoot to get the other one is the avoidable cost here.

If you're working through a large generated set, the same triage applies with a couple of extra checks specific to synthetic images — we walk through the full process in how to choose the best AI headshot from 50 results.

The short version of all of it: smile, but not as an instruction to your mouth. Give yourself something real to react to, reset between frames, and judge the result at the size people will actually see it. A moderate, genuinely engaged expression is the safest answer in nearly every field — and unlike almost everything else in a headshot, it costs nothing but a few more shutter clicks to get right.