How to Stop Crying

By Blubee Editorial, Editorial team · Updated August 10, 2026

On this page

Crying is a two-stage physiological event, not a failure of control: sympathetic arousal builds first, tears arrive near the peak, and calming activity follows behind. That is why an episode responds to physical interruption rather than to willpower. Immersing the face in 10°C water while holding the breath drops heart rate by 22% within ten seconds.

This page covers what the body is doing during an episode, eight techniques that interrupt one, the difference between interrupting tears and suppressing the feeling, what crying actually does to mood, and the cases where stopping is the wrong goal.

What your body is doing during a crying episode#

A crying episode is an arousal curve rather than a steady state, and where you are on that curve decides which technique will work. Distress builds first as sympathetic activation — faster heart rate, rising skin conductance, shallow and irregular breathing — and tears appear close to the peak. Parasympathetic activity, the calming branch, comes in behind them.

The timing matters more than the labels do. In physiological recordings the calming effects of an episode, such as slowed breathing, outlast the arousing effects, such as raised heart rate, by two to three minutes. That lag is why researchers describe crying as partly self-soothing, and it is the closest measurement there is to feeling wrung out but steadier. It is not evidence that an episode ends below the arousal level it started from.

The lump in the throat has a mechanism too. It comes from muscles around the glottis contracting while you hold breath and voice steady, which is why swallowing and yawning loosen it while clenching the jaw does not.

Tears arrive near the top of the arousal curve, not at the bottom. Anything that lowers arousal shortens the episode; anything that raises it extends the episode instead.

Every technique below acts on arousal. None of them acts on the feeling underneath, and none of them is supposed to.

The 8 techniques that actually interrupt an episode#

Techniques that work on crying are physical, fast and aimed at the nervous system rather than at the thought that started it. The first two have the clearest physiological evidence behind them and are the ones to reach for when there are seconds rather than minutes. The rest are ordered by how quickly they take effect.

1. Cold water on the face#

Cold across the forehead, eyes and cheeks triggers the diving response, a reflex that slows the heart within seconds. In one physiological study, immersing the face in 10°C (50°F) water while holding the breath produced a 22% fall in heart rate that developed within ten seconds. A wet paper towel, a cold can held against the cheekbones, or half a minute at a cold tap gets a smaller version of the same effect.

This is the fastest available interruption and the most usable in public, since a trip to a bathroom sink is unremarkable. It fits when arousal has to come down inside a minute. It does not fit where a heart condition or medication affecting heart rate is in the picture, since a deliberately triggered slowing of the heart is worth raising with a doctor first.

2. Slow the breath to about six a minute#

Breathing rate determines how much brake the vagus nerve applies, and one rate does it best. Heart rate oscillations reach their maximum amplitude at roughly 0.1 Hz — about six breaths a minute — the point at which breathing and the baroreflex fall into phase. In practice that is a four-second inhale and a six-second exhale, counted rather than estimated.

The counting is doing half the work, because it occupies the same attention that would otherwise keep running the thought feeding the episode. Use this when two or three uninterrupted minutes are available. Do not turn it into breath-holding: holding the breath raises arousal instead of lowering it, which is the opposite of the intended effect.

3. Loosen the throat instead of clenching the jaw#

The lump in the throat is muscular, and fighting it tightens it further. Swallowing deliberately two or three times, yawning, or letting the jaw hang open a centimetre releases the constriction around the glottis that produces the sensation. Saying one full sentence out loud does the same thing, by putting the vocal folds back into ordinary use.

Clenching the jaw and pressing the tongue hard against the roof of the mouth — the version circulated most widely online — works in the opposite direction. It adds muscular tension in exactly the region that is already contracting. It may delay the first tear by a few seconds, at the cost of a harder episode when the delay runs out.

4. Cut the input before you work on the feeling#

Crying escalates with continued input, so removing the input is usually faster than any regulation technique. Leaving the room, turning the phone face down, stopping a message thread mid-sentence or simply turning away from the other person’s face cuts the stream of signal driving the episode.

This is not avoidance if you set a return. Saying the sentence out loud — back in ten minutes — makes the exit read as a pause rather than a walkout, to you and to the other person. It fits arguments, meetings and anything happening on a screen. It does not fit a conversation where leaving would land as abandonment on someone who is also distressed.

5. Change what your body is doing#

Posture and movement change respiration faster than any instruction to calm down. Standing up, walking to another room, carrying something heavy for thirty seconds, or running cold water over the wrists all shift breathing pattern and muscle tone without requiring a decision about the feeling.

The reason this belongs on the list is narrow, and it is not that movement is generally good for you. It works because it changes the mechanical input to the respiratory and cardiovascular systems while the arousal curve is still climbing. Two minutes is enough. Anything longer is a different intervention with different aims.

6. Say the trigger in one sentence#

Naming what set the episode off, out loud or in writing, turns an undifferentiated surge into a specific event with edges. One sentence is the format: this is about being talked over in that meeting, not a paragraph of analysis about why it happened.

Precision does the work here. A vague label keeps the episode general and therefore large, while a specific one makes it finite and time-stamped. If nothing comes, the honest sentence is that you do not know yet what this is about, which is still more contained than silence. Skip this one mid-argument, where naming a trigger out loud converts straight into an accusation.

7. Give it five minutes instead of a ban#

A deliberate time box works where a prohibition does not, because a prohibition stacks a second stressor on top of the first. Set five minutes, let the episode run without commentary, then move. In most episodes the arousal curve is already falling by then, and the wrung-out steadiness that follows is easier to work with than a half-suppressed surge.

This applies when you are alone and the next thing in the day can wait. It does not apply thirty seconds before a presentation, which is precisely the situation the first two techniques exist for.

8. Let someone be there, if that is available#

Company changes the outcome of an episode, and the diary evidence is more specific about how than any technique list is. The episodes that improved mood were largely the ones with support present, or the ones that ended in some resolution or new understanding. Episodes with neither mostly changed nothing, and crying in front of two or more people went with worse or unchanged mood rather than better.

That is an uncomfortable finding when reaching out feels expensive. Plenty of people learned early that visible distress had a price, and they worry that asking will read as too much — the label itself is worth examining, and what clingy actually means is narrower than its use suggests. For a structured read on how you handle closeness under stress, a 12-item questionnaire lets you see where your answers land on anxiety and avoidance, free and without an email.

Interrupting the episode is not suppressing the feeling#

Interrupting tears and suppressing an emotion are two different operations with two different costs, and this page is only about the first. Interrupting changes arousal and timing. Suppressing changes the outward signal while the internal event carries on, and experimental work on expressive suppression keeps finding the same split: visible expression drops while sympathetic markers rise.

The distinction matters because the goal is almost always timing rather than deletion. Choosing not to cry in a performance review and then crying in the car is regulation working correctly. Choosing never to cry, and treating each episode as a failure, is the version that produces the internal cost without the benefit.

People who grew up where every feeling was treated as everyone’s business often read any interruption as dishonesty, because privacy was never a category — family closeness with no boundary describes that setup. The correction is not more tears or fewer; it is the recognition that timing was always yours to set.

Crying is also not a symptom by default. Across a five-country sample of 893 adults, women reported crying 4.6 times a month and men 1.5, which is a routine event rather than a rare one.

What crying actually does to your mood#

The belief that a good cry reliably helps is only sometimes true, and the moment you check is what decides the answer. In a daily diary study of 1,004 crying episodes recorded by 97 women, about 30% were followed by improved mood, 60% by no change, and 9% by worse mood. That is a much weaker result than the folk version claims.

The laboratory picture explains the gap. In a quasi-experimental study, people who cried at a film reported worse mood immediately afterwards, had returned to their pre-film baseline by 20 minutes, and reported less negative mood than baseline at 90 minutes. Non-criers showed no such pattern.

Immediately after: worse. At 20 minutes: back to where you started. At 90 minutes: better than before. Judging an episode in its first ten minutes gives the worst available reading.

The practical consequence is about measurement, not about crying. If you decide whether an episode helped while your face is still hot, the answer will be no almost every time, and the conclusion you draw about yourself will be built on the least informative data point available. Noticing the pattern across days rather than minutes is the whole reason apps built for daily tracking exist.

Who you cry in front of changes the outcome#

Who is in the room is one of the few conditions that reliably separates a crying episode that helps from one that changes nothing. That raises a question the techniques above cannot answer: whether reaching out for company during distress is even available to you as an option.

Attachment research describes this with two continuous dimensions rather than a set of types. One is anxiety about losing connection; the other is discomfort with depending on anyone. Most people sit somewhere in the middle of both, and the four labels drawn from those axes — described in full across the four broad patterns — are corners of a map, not categories people fall into cleanly.

The two ends produce opposite crying behaviour. Answers toward the anxious end of the scale often go with crying that escalates when the other person does not respond, because the tears are partly a signal that contact is needed. Answers toward the other end go with handling distress alone, leaving the room first and reporting afterwards that it was nothing.

Neither position is a defect, and neither is a diagnosis. Knowing which way your answers lean mostly tells you which of the eight techniques to try first: cutting the input works well for one pattern and badly for the other.

When stopping crying is not the explanation#

Sometimes the tears are not the problem to solve, and treating them as one delays the thing that actually needs attention. Four situations sit outside everything above, and none of them is fixed by cold water.

Grief is the first. Frequent crying after a loss is the process doing what it does, and the frequency drops on its own timeline rather than on a schedule you set. Applying interruption techniques to grief works fine for getting through a workday and does nothing to the underlying course.

The second is crying with no trigger and no relief. Episodes that do not match the situation, or laughing and crying that arrive without a matching feeling, can point at pseudobulbar affect — a neurological condition estimated to affect about 2 million people in the US, in which crying is decoupled from mood. It appears alongside conditions such as stroke, multiple sclerosis and traumatic brain injury, and it responds to medical treatment, not to breathing exercises.

The third is physical and pharmacological. Thyroid problems, a recent medication change, chronic pain and long-term sleep deprivation all lower the crying threshold, and every one of them is a question for a doctor rather than a technique to practise.

The fourth is duration. Crying most days for weeks, alongside persistent low mood or loss of interest, is a pattern to bring to a clinician. No questionnaire settles it, and the free tests on this site, listed at the tests hub, describe patterns rather than conditions. What a questionnaire can say is how you tend to handle distress. What it cannot say is why this week is harder than the last one.