The image of pouring black coffee into someone who has had too much to drink is a staple of films and family gatherings. Search suggestions show people asking whether it is bad to drink coffee after alcohol, and what happens if you drink both together. The science is clear on the central point: coffee does not remove alcohol from the body. What caffeine does do is change how drunk you feel and how some aspects of your performance look, and that mismatch is where the real risk lies.
What sobering up actually means
Blood alcohol concentration falls as the liver metabolises alcohol, at a rate that nothing in a cup of coffee meaningfully speeds up. Being sober means that level has fallen and the impairment it caused has gone. Feeling more awake is not the same thing. Caffeine is a stimulant and alcohol is a depressant, so the question is really whether caffeine cancels alcohol's effects on the brain and body, and if so, which ones.
Caffeine does not change blood alcohol
Ferreira and colleagues, in Alcoholism: Clinical and Experimental Research[1], tested this directly with 26 young healthy volunteers in two groups receiving either a moderate or a higher dose of alcohol as vodka. Each person completed three sessions a week apart, in random order: alcohol alone, an energy drink alone, and alcohol with an energy drink. The researchers measured breath alcohol concentration, subjective feelings of intoxication, objective motor coordination and visual reaction time.
The energy drink did not alter breath alcohol concentration in either group. Compared with alcohol alone, alcohol plus the energy drink significantly reduced people's perception of headache, weakness, dry mouth and impaired motor coordination. But the energy drink did not significantly reduce the actual deficits alcohol caused in objective motor coordination and visual reaction time. In other words, people felt less impaired than they were.
Driving: partly better, still impaired
Driving is the most consequential real-world test. Liguori and Robinson, in Drug and Alcohol Dependence[2], gave 15 adults a capsule with 0, 200 or 400 mg of caffeine followed by a drink containing either no alcohol or enough alcohol to approach the US legal limit of 0.08% breath alcohol, in a double-blind randomised design. Forty-five minutes later, they tested subjective effects, balance, reaction time, attention and simulated driving.
Alcohol alone made people feel dizzy and high, slowed choice reaction time and braking, and increased body sway. Both caffeine doses counteracted alcohol's slowing of braking speed to a similar degree, but they did not correct alcohol's effects on choice reaction time or body sway. And even with caffeine, braking remained significantly slower than with placebo. The authors concluded that caffeine may increase alertness and improve reaction time after alcohol, but will not completely counteract alcohol impairment in a driver.
Faster, but not more in control
Marczinski and Fillmore[3] looked at a more subtle aspect of impairment: behavioural control. They tested 12 social drinkers on a task requiring quick responses to some targets and sudden suppression of responses to others, under three doses of caffeine combined with either no alcohol or a moderate dose of alcohol. Alcohol impaired both the speed of responding and the ability to hold back a response.
Caffeine counteracted alcohol's effect on response speed, but had no effect on alcohol's impairment of inhibitory control. That distinction matters. Inhibitory control is what stops you doing something you should not, such as pulling out into traffic, sending a message you will regret, or accepting another drink. Caffeine appeared to restore the go without restoring the stop.
Feeling stimulated while still impaired
Two further experiments with energy drinks, by the same research group, filled in the subjective side of this picture. In the first, Marczinski and colleagues[4] randomly assigned 56 participants to alcohol, an energy drink, alcohol mixed with an energy drink, or a placebo drink. Alcohol alone impaired both response speed and the ability to hold back responses. Adding the energy drink counteracted some of the slowing, but not the loss of inhibition, which repeated the earlier caffeine finding. On subjective measures, the energy drink increased feelings of stimulation, while ratings of impairment and intoxication were similar to alcohol alone. The authors described the mix of impaired inhibition and enhanced stimulation as potentially riskier than alcohol alone.
In the second, Marczinski and colleagues[5] tested 18 men and women in four sessions. Alcohol slowed information processing when doing two things at once and impaired both simple and complex motor coordination, and adding the energy drink did not change any of those impairments. But the combination reduced people's feelings of mental fatigue and increased their feelings of stimulation compared with alcohol alone. The authors warned that drinkers may wrongly perceive themselves as better able to function than they are.
Larger doses of alcohol, no benefit at all
At higher levels of intoxication, even the partial benefits seen in some studies may disappear. Howland and colleagues, in Addiction[6], randomised 127 young adult heavy episodic drinkers to beer or non-alcoholic beer, with or without added caffeine at the level found in a commercial caffeinated beer, targeting a breath alcohol level of 0.12%. Alcohol significantly impaired driving performance and sustained attention and reaction time, with mostly large effects. The addition of caffeine had no main or interaction effects on performance. The authors concluded that adding caffeine to alcohol does not appear to improve driving or attention relative to alcohol alone.
Believing coffee works can make things worse
Perhaps the most important experiment in this area is about expectation. Fillmore, Roach and Rice, in the Journal of Studies on Alcohol[7], gave 42 social drinkers a moderate dose of alcohol with either caffeine or a placebo caffeine, and told some groups that caffeine would counteract alcohol's impairing effects and other groups that it would have no effect. They then tested psychomotor performance over three hours.
Groups led to expect that caffeine would counteract alcohol showed greater impairment than those led to expect no counteraction. Caffeine itself had no significant effect on alcohol impairment in this study. The authors' explanation is that drinkers who expect to be impaired try harder to compensate, while drinkers who believe something has protected them stop compensating, and so perform worse. The belief that coffee sobers you up may therefore make people less careful at exactly the wrong moment.
Real-world risk: energy drinks and alcohol
Laboratory studies cannot ethically recreate heavy real-world intoxication, so survey data provide a useful complement. O'Brien and colleagues, in Academic Emergency Medicine[8], surveyed a stratified random sample of 4,271 college students at 10 universities in the United States. Of students who had drunk alcohol in the past month, 24% had mixed alcohol with energy drinks.
In multivariable analyses, those students reported more heavy episodic drinking, 6.4 days compared with 3.4 days on average, and twice as many weekly episodes of drunkenness. They also had significantly higher rates of alcohol-related harms, including being taken advantage of sexually, riding with an intoxicated driver, being physically hurt or injured, and needing medical treatment, even after adjusting for the amount of alcohol consumed. This is a cross-sectional survey, so it cannot show that energy drinks caused these harms, since people who seek out those mixtures may already be more inclined to risk. But it is consistent with the laboratory picture of people feeling less drunk than they are.
Why caffeine and alcohol interact this way
A review by Ferré and O'Brien[9] proposed a neurochemical explanation. Both substances act on adenosine signalling in the brain. Caffeine blocks adenosine A1 receptors, which appear to mediate some of alcohol's sedating and unsteadying effects, so caffeine may suppress the unwanted signals that normally tell a drinker to stop. At the same time, blocking A2A receptors may amplify the dopamine-related rewarding effects of alcohol. The authors described the combination as a perfect storm that could increase drinking and alcohol-related consequences. This is a proposed mechanism drawn from the literature, and it has not been fully confirmed in humans.
Does coffee help a hangover?
The next-morning question is separate. Rohsenow and colleagues[10] randomised university students and recent graduates to beer with or without added caffeine, to a breath alcohol level of 0.12%, followed by an eight-hour supervised sleep period. Caffeinated beer improved perceived sleep quality and morning alertness somewhat, but had no effect on hangover incidence or severity, and no effect on how long it took to fall asleep or on total sleep time.
That study tested caffeine during drinking rather than coffee the next morning, and we did not find strong trials of morning coffee as a hangover remedy. A morning coffee may help a regular drinker avoid adding caffeine withdrawal headache to a hangover headache, and it may improve alertness, but there is no good evidence that it treats a hangover.
Why the myth has lasted so long
Given how consistent this evidence is, it is fair to ask why so many people still believe coffee sobers you up. The studies above suggest an answer. The effects caffeine does have are exactly the ones a drinker notices: less tiredness, more stimulation, fewer of the sleepy, heavy feelings of intoxication, and in some tests a slightly quicker reaction. The effects it does not have are the ones a drinker cannot easily notice in themselves: slower information processing, poorer coordination, weaker impulse control, and an unchanged blood alcohol level. A person who feels more awake after a coffee has every reason, from their own experience, to believe it worked.
That is also why the expectancy result matters so much. If the belief itself makes people stop compensating for their impairment, then the myth is not harmless folklore. It actively changes behaviour in the direction of more risk.
A note on coffee, alcohol and the liver
One genuinely interesting finding sits at the other end of the timescale. Klatsky and colleagues[11] followed 125,580 members of a US health plan without known liver disease, 330 of whom later developed cirrhosis. Compared with non-drinkers, people drinking four or more cups of coffee a day had a relative risk of alcoholic cirrhosis of 0.2, and the inverse association with liver enzymes was stronger among heavy alcohol drinkers. Tea was unrelated. This is observational and says nothing about coffee's effects on a night of drinking, but it is part of a wider body of research on coffee and liver health, covered in our article on coffee and the liver. It is not a reason to drink alcohol, and it is not protection against the acute harms described above.
The limits of this evidence
Most of the controlled experiments are small, and ethical limits mean they test moderate levels of intoxication, not the very high levels reached in real-world binge drinking. Several use caffeine capsules or energy drinks rather than coffee, although the caffeine is the same active ingredient. The survey of students is cross-sectional and from one country. The neurochemical explanation is a hypothesis. None of this evidence measures what happens when someone drinks heavily and then drinks coffee before driving, which is the scenario that matters most and cannot be tested ethically.
The bottom line
Coffee does not sober you up. It does not lower blood alcohol, and while caffeine can make you feel less drunk and partly speed up some responses, controlled studies found driving still impaired, inhibitory control untouched, and no benefit at higher levels of intoxication. People who believed caffeine would protect them actually performed worse, and people who mix alcohol with energy drinks report more drunkenness and more harm.
If you like a coffee at the end of a night out, that is not dangerous in itself for most healthy adults, as long as it is understood for what it is: a drink that will make you feel more alert, possibly keep you awake later, and do nothing to the alcohol already in your blood. The problems arise when it is used as a reason to keep drinking, to drive, or to make a decision that alcohol has already impaired.
The only thing that sobers you up is time. If you have been drinking, coffee is not a reason to drive, and a wide-awake feeling is not evidence that you are safe. If you want to understand how long caffeine itself stays in your system, including into the night after an evening coffee, see our article on how long caffeine lasts. Nothing here is medical advice.