This is not medical advice. General information about food tracking, not a treatment plan, and not advice about drinking. If you are under 18, pregnant or breastfeeding, have ever had an eating disorder or a difficult relationship with food, are managing a medical condition, or are concerned about your own alcohol use, talk to a doctor or a registered dietitian before you start counting calories or change your intake. The full list is here.
I have been building a calorie counter for two years and the thing it is worst at is a drink. Not because the arithmetic is hard. Because almost every input the app needs is missing by design: there is no number on the bottle, no macro to file it under, and no reliable way to tell how much was in the glass.
This post is about the calories in alcohol and how to get them into a log. It is not about whether to drink, which is a separate question and not one a calorie app has any standing to answer. If you are worried about your own drinking, that is worth raising with a doctor rather than with an app.
Nobody prints the calories on a bottle, and that is legal
Pick up a tub of yoghurt in Britain and the energy, fat, carbohydrate, sugar, protein and salt are on the back in a standard grid. Pick up a bottle of wine and there is an alcohol percentage, a volume, and a line about sulphites.
That gap is written into the law. Article 16(4) of the retained Regulation 1169/2011 says the list of ingredients and the nutrition declaration "shall not be mandatory for beverages containing more than 1,2 % by volume of alcohol".2 Two of the fourteen particulars every other food has to carry are simply switched off above 1.2% strength.
The American position is different in mechanism and identical in effect. Alcohol labelling sits with the Alcohol and Tobacco Tax and Trade Bureau rather than the FDA, and TTB's standing guidance since May 2013 has been that Serving Facts panels carrying calories and macronutrients are authorised on a voluntary basis, pending the completion of rulemaking.3 Some brands print one. Nothing requires any of them to.
So a drinker is guessing, and a review of eighteen studies found they guess badly: pooling the eight that tested it, 74% of participants were inaccurate about the energy content of alcoholic drinks (95% CI 64% to 82%), and 64% supported mandatory energy labelling. The same review looked at six studies of whether labelling changes what people drink and found no benefit, while grading its own evidence on that point as very low quality, since none of the six ran in a real bar or shop.4
That second finding is worth sitting with before assuming a label would have fixed anything. Knowing the number and acting on it are different things, and only the first is what this post can help with.
What a drink actually costs
USDA FoodData Central does carry alcoholic drinks, and its Standard Reference values are the closest thing to a neutral reference here.5 The table below is those figures converted to the volumes people are actually served, using USDA's own gram weight per fluid ounce so nothing depends on a density I made up.
| Drink | Serve | Calories | Alcohol |
|---|---|---|---|
| Beer, regular (5.0%) | Pint, 568 ml | 245 | 22.2 g |
| Beer, regular (5.0%) | Bottle, 330 ml | 143 | 12.9 g |
| Beer, regular (5.0%) | Can, 12 fl oz | 153 | 13.9 g |
| Red wine (13.4%) | Small glass, 125 ml | 106 | 13.2 g |
| Red wine (13.4%) | Standard glass, 175 ml | 148 | 18.4 g |
| Red wine (13.4%) | Large glass, 250 ml | 211 | 26.3 g |
| White wine (13.0%) | Standard glass, 175 ml | 143 | 17.9 g |
| Spirits (39.8%) | Single, 25 ml | 54 | 7.8 g |
| Spirits (39.8%) | Double, 50 ml | 109 | 15.7 g |
| Spirits (39.8%) | Jigger, 1.5 fl oz | 97 | 14.0 g |
The strengths in the left column are not decoration. They are what USDA's entries work out to once you convert grams of ethanol per 100 g back into alcohol by volume, and your drink will not match them. A 4% session bitter and an 8% IPA land on the same row in a generic calorie chart and are not the same drink.
Three pints of that 5% beer is about 736 calories. For a woman on a 1,600 calorie target that is 46% of the day, and nothing in the evening said so out loud.
Spirits come out looking cheap, and they are, right up until the mixer. A single measure is 54 calories and the tonic it arrives in can be more than that again. The spirit is the part the table covers; the mixer is an ordinary soft drink and belongs in the log as one.
Alcohol is not a macronutrient, so a macro tracker has nowhere to put it
Alcohol carries energy. FAO's conversion factors give it a rounded Atwater value of 29 kJ/g, which is 7.0 kcal per gram, against 4 for protein, 4 for carbohydrate and 9 for fat.1 It is the fourth energy-bearing thing people consume in quantity. It is also not one of the three fields a macro tracker has.
Which leaves every calorie logger with the same choice, and no good option:
- Log the drink as calories with no macros. The daily calorie total is right and the macro rings never add up to it, which looks like a bug and gets people filing bug reports.
- Convert the calories into carbohydrate at 4 kcal per gram. The totals reconcile and the carbohydrate figure is now fiction: 148 calories of wine becomes 37 g of carbohydrate, where the actual carbohydrate in that glass is about 4.5 g.
- Split it between carbohydrate and fat. Same fiction, spread over two numbers.
Plenty of advice online recommends the second or third of those without saying what it costs. It costs the meaning of your macro numbers on any day you drink. If you are following a protein target, the calorie side of the ledger is what alcohol belongs in, and the honest answer to "which macro" is none of them.
My own preference, and it is a preference rather than a finding: log the drink for its calories, leave the macros at zero, and accept that the rings will not close on a Friday. A macro total that is 300 calories short and correct is more useful than one that reconciles and lies.
The pour is a bigger error than the drink
Everything above assumes you know how much was in the glass. In a bar you do not, and the size of that error has been measured.
A research team bought and measured drinks in 80 establishments across ten Northern Californian counties, using the brand or a laboratory analysis of a sample to get the alcohol concentration. The average wine drink contained 43% more alcohol than a US standard drink, with no difference between red and white. The average draught beer was 22% over. Spirits split by type: a straight shot came out at about one standard drink, while the average mixed drink was 42% over. Variation was widest for wine and for mixed spirit drinks.6
A US standard drink is 14 g of ethanol, which at 7 kcal per gram is about 99 calories before anything else in the glass. Forty-three percent over makes it 142. Log two glasses of wine at a textbook 125 ml and the real pour may have been closer to a large glass each, and you are roughly 200 calories light on the day without having made a single mistake in the app.
Two caveats the authors would want attached. This was Northern California in the 2000s, not a universal law of bartending, and a UK pub pouring a metered 175 ml is a different setting with a legal measure behind it. And they measured alcohol content rather than calories, so it transfers cleanly to wine and spirits, where ethanol is nearly all of the energy, and less cleanly to a sweet mixed drink where the syrup is doing its own work.
This is the same failure that makes meals you did not cook hard to log, arriving through a different door. There the unknown is what went in the pan; here it is what went in the glass.
What drinking does to the food you log around it
The drink is the part you can at least estimate. The part that does not appear in any log is what happens to dinner.
A 2019 systematic review searched four databases through February 2018 for crossover and randomised trials comparing an alcohol dose against a non-alcoholic condition, and found 22 studies covering 701 participants. They pointed consistently the same way: no compensation. People did not eat less to make room for the energy in the drink. Meta-analysing the twelve studies that allowed it, alcohol raised food energy intake by a weighted mean of 343 kJ, which is about 82 calories (95% CI 161 to 525 kJ), and total energy intake by 1,072 kJ, about 256 calories (95% CI 820 to 1,323 kJ).7
The authors were direct about where that does and does not apply. Generalisability is limited to younger adults, aged 18 to 37, and some of the pooled outcomes carried substantial statistical heterogeneity or signs of small-study effects.7 It is a laboratory result about single occasions, not a life.
Still, it reframes what a drink costs. The 245 calories in a pint is the part you can see. If the review's average holds for you, there is another 82 or so on the plate beside it, and that part arrives as food you will log accurately and never attribute to the drink.
Do you get all seven calories?
This is the part of the subject where the honest answer is more interesting than the confident one, and where you will find plenty of people online claiming alcohol calories do not count.
They partly do not, and the size of the discount is known within a range rather than exactly.
Ethanol has a large thermic effect. Six healthy non-alcoholic adults drank 95.6 g of ethanol across three meals in a respiration chamber and raised 24-hour energy expenditure by 5.5% against a control day, which the authors calculated as ethanol-induced thermogenesis of 22.5% of the energy ingested. In the fasting state, 31.9 g of ethanol raised expenditure 7.4% over baseline, for a calculated 17.1%. They concluded that roughly 80% of ethanol energy is available as metabolizable energy in healthy moderate drinkers.8
Six subjects, twice. It is a small study and the two conditions disagreed by five percentage points between them.
FAO makes a smaller correction on the same grounds. Alongside the 7.0 kcal/g metabolizable energy factor it lists a net metabolizable energy factor of 6.3 kcal/g for alcohol, a difference of 10%, which it attributes to thermogenesis following alcohol consumption.1 Food labels worldwide use the 7.0 figure, so nothing you read on a packet has the correction in it.
Set the two effects against each other on one standard drink. The thermogenic discount is somewhere between 10 and 20 calories out of 99. The extra food eaten alongside, at the review's average, is about 82.7 Those are different kinds of number and I would not subtract one from the other on a spreadsheet, but they point in opposite directions and the food side is several times larger at ordinary drinking volumes.
Practical upshot: log alcohol at the 7 kcal per gram the labels use. The real discount is somewhere between 10% and 20%, it is smaller than the pour error in any bar, and it is much smaller than the food that turns up alongside the drink.
What the evidence does not settle
Whether drinking makes people gain weight is not a settled question, and anyone telling you otherwise is ahead of the literature.
The case that it does starts with a metabolic chamber study from 1992. Eight normal men spent two 48-hour sessions in indirect calorimetry. In one, ethanol was added on top of their requirement at 96 g a day; in the other, ethanol replaced 25% of their energy requirement, substituted isocalorically for fat and carbohydrate. Both lowered 24-hour lipid oxidation, by 49.4 and 44.1 g a day, reductions of 36% and 31% from the control day. The effect appeared only during the daytime window when the ethanol was being metabolised, and carbohydrate and protein oxidation were unchanged. The authors concluded that habitual consumption in excess of energy needs "probably favors lipid storage and weight gain".9
The verb is doing a lot of work there, and so is the condition attached to it: probably, and in excess of energy needs.
One of the same authors reviewed the field eight years later and drew a narrower line: ethanol cannot be stored, so it becomes the priority fuel and displaces fat oxidation while it is being cleared, but there is no evidence that ethanol consumed under normoenergetic conditions, meaning isoenergetically replacing carbohydrate or fat, leads to greater body fat storage than either. The same review flagged an open gap, that nobody had properly studied what alcohol does to spontaneous physical activity.10
And the epidemiology does not close it either. A systematic review of 31 publications from 1984 to 2010 found that large cross-sectional studies and well-powered prospective cohorts with long follow-up were contradictory, and that short-term experimental trials showed no clear trend. Its overall conclusion was that the results "do not conclusively confirm a positive association between alcohol consumption and weight gain", with positive findings turning up mainly in studies of heavier drinking.11
That review also reported that light to moderate intake, wine in particular, may be more likely to protect against weight gain, while spirits have been positively associated with it.11 I am quoting it because leaving it out would be selective, and I would not act on it. It is observational, the authors themselves called the body of evidence contradictory, and who drinks wine moderately differs from who drinks spirits heavily in a hundred ways a questionnaire cannot adjust for.
What survives all of that is narrow and useful. Alcohol carries energy, people do not eat less to make room for it, and the energy is invisible unless you write it down. Whether it does something to your body beyond the arithmetic is not established.
How to log a drink when calorie tracking has no entry for it
The method is two numbers you can get off the bottle and one multiplication.
Millilitres times strength gives you millilitres of ethanol. Ethanol weighs 0.789 g per millilitre. Multiply the grams by 7 and you have the calories from alcohol.1
A 175 ml glass of 13% wine: 175 times 0.13 is 22.75 ml of ethanol, times 0.789 is 17.9 g, times 7 is 126 calories.
Then add what else is in it. Red wine carries about 2.6 g of carbohydrate per 100 g,5 so that glass has roughly 4.5 g more, worth 18 calories. Total 144. USDA's own entry for table red gives 148 for the same volume,5 but its wine works out to 13.4% rather than 13%, which accounts for most of the four calorie gap between the two routes.
Four rules that matter more than the decimal places:
- Log the drink before you drink it, not at the end of the night. This is the single biggest improvement available and it has nothing to do with accuracy.
- Round the pour up rather than down. The measured bias in bars runs 22% to 43% over a standard drink, and it runs one way.6
- Log the mixer separately. It is an ordinary soft drink and your app already handles those.
- Read the week, not the night. A drinking evening will blow a daily target and tell you nothing; seven days including it is a number you can act on.
Save one entry per drink you have regularly and reuse it. The precision you lose by not recomputing a familiar drink each time is far smaller than the precision you lose by giving up on logging drinks at all, which is what most people do. Consistency beating accuracy is the general rule for tracker accuracy and it applies here more than anywhere.
What a photo scan can see here, and what it cannot
MyPlate is a photo-first calorie counter, so it is worth being plain about where that approach falls down, and drinks are where it falls down hardest.
A photo of a glass of red wine shows a glass of red wine. It does not show whether the wine is 11% or 15%, which is a 36% swing in the alcohol calories. It cannot tell a 125 ml pour from a 250 ml one without a reference object, and wine glasses vary more in size than dinner plates do. A gin and tonic and a tonic with a lime look identical from above.
None of that is fixable by a better model. The information is not in the picture.
So for drinks I use the text path rather than the camera, and type what the photo cannot carry: "175 ml glass of red wine, 13%". MyPlate has no food database and no barcode scanner, so there is no brand entry to look up, which is a real limitation if you drink something unusual. What the description does give you is the two numbers that actually determine the answer.
The same honesty applies to the rest of the log. A scan is an estimate, not a measurement, and it is a worse estimate for a drink than for a plate of food.
Common questions
How many calories are in a glass of wine?
About 148 calories in a 175 ml standard glass of red wine, 106 in a 125 ml small glass and 211 in a 250 ml large one, using USDA FoodData Central values for table red, which works out to a 13.4% wine.5 White wine at 13.0% is a few calories lower for the same volume, at 143 for a 175 ml glass. Strength is what moves the number most, so an 11% wine and a 15% wine differ by roughly a third in the same glass. If you want it exactly, multiply millilitres by the strength on the bottle, multiply by 0.789 to get grams of ethanol, then by 7, and add about 4.5 g of carbohydrate per 175 ml glass.
Should I log alcohol as carbs or as fat?
As neither, if your app will let you. Alcohol is a fourth energy source at about 7 calories per gram and it is not protein, carbohydrate or fat.1 Converting a drink's calories into carbohydrate at 4 calories per gram makes the daily totals reconcile, but it turns a 148 calorie glass of wine into 37 g of carbohydrate when the real figure is about 4.5 g, and any protein or carbohydrate target you are following is then reading a number that is mostly invented. Log the calories, leave the macros at zero, and accept that the macro rings will not add up to the calorie total on a day you drink.
Do alcohol calories count the same as food calories?
Not quite, and the discount is smaller than the internet suggests. Ethanol has a large thermic effect: in a respiration chamber study of six healthy adults, ethanol-induced thermogenesis came to 22.5% of the energy ingested when drunk with meals and 17.1% when drunk fasted, leading the authors to conclude that roughly 80% of ethanol energy is available as metabolizable energy.8 FAO applies a smaller correction, listing a net metabolizable energy factor of 6.3 kcal/g against the 7.0 used on labels, a 10% difference it attributes to the same thermogenesis.1 So somewhere between 10% and 20% of a drink's alcohol calories go out as heat. That is less than the error in a typical bar pour, and much less than the extra food people eat alongside a drink, which averaged about 82 calories across twelve randomised and crossover trials.7 Log the full 7 per gram.
Why isn't my drink in my calorie counter app?
Because the manufacturer was never required to work the number out. Under Article 16(4) of the retained Regulation 1169/2011, the ingredients list and the nutrition declaration are not mandatory for any beverage over 1.2% alcohol by volume,2 and in the United States the Alcohol and Tobacco Tax and Trade Bureau has allowed Serving Facts panels on a voluntary basis since 2013 rather than requiring them.3 A database can only carry a figure somebody worked out and published, so the gap on the bottle tends to become a gap in the app. The workaround is to calculate it from the volume and the strength, which are the two things a bottle does have to tell you.
What should I look for in a calorie tracking app if I drink regularly?
Whether it lets you save a custom entry and reuse it in one tap, and whether it shows you a weekly figure as well as a daily one. Those two features matter more than database size here, because no database can know the strength of your wine or the size of your pour, and a drinking evening read as a single blown day tells you nothing useful. Being able to enter calories without being forced to assign macros is worth having too, since alcohol genuinely belongs to none of the three. For the record, MyPlate has no food database and no barcode scanner, which is a real limitation for packaged drinks; what it does have is a text description path, so "175 ml glass of red wine, 13%" gets further than a photograph of the glass will.
Log the drink first
Alcohol is the one thing most people eat or drink that carries real energy and arrives with no number attached. The law says the bottle does not have to tell you, the macro fields in your app have nowhere to put it, and the pour in front of you is probably bigger than the one in the reference table.
What you can do about that is small and it works. Get the volume and the strength, multiply them out, log it before you drink rather than after, and read the week rather than the night. A pint is about 245 calories and a large glass of wine is about 211,5 and once those are in the log the rest of the week stops being a mystery.
One thing not to do: do not cut food to make room for the drinks in advance. If you are eating in a deficit, the floor worth respecting is 1,200 calories a day for women and 1,500 for men, which is where the NIH clinical guidelines put it and where MyPlate's own targets stop regardless of what the formula produces.12 Skipping dinner to bank calories for a night out is how a deficit turns into a day of not eating.
And if working out the calories in a drink has turned an evening with friends into a problem to be solved, that is a signal to put the app down rather than to log more carefully. Tracking that has become something to get right, rather than something that helps, is worth raising with a doctor or a registered dietitian.
Sources
Every figure above traces to one of these. If you find a number that doesn't match the source it claims, tell me and I'll correct it.
- Food and Agriculture Organization of the United Nations (2003). Food energy: methods of analysis and conversion factors. FAO Food and Nutrition Paper 77, Rome. Read on FAO →
- Regulation (EU) No 1169/2011 on the provision of food information to consumers, Article 16(4). Assimilated direct legislation, United Kingdom. Read on legislation.gov.uk →
- Alcohol and Tobacco Tax and Trade Bureau (2013). TTB Ruling 2013-2: Voluntary Nutrient Content Statements in the Labeling and Advertising of Wines, Distilled Spirits, and Malt Beverages. US Department of the Treasury. Read the ruling →
- Robinson E, Humphreys G, Jones A (2021). Alcohol, calories, and obesity: a rapid systematic review and meta-analysis of consumer knowledge, support, and behavioral effects of energy labeling on alcoholic drinks. Obesity Reviews, 22(6):e13198. Read on PubMed →
- US Department of Agriculture, Agricultural Research Service (2018). FoodData Central, Standard Reference Legacy Release, April 2018. USDA National Agricultural Library. Read on FoodData Central →
- Kerr WC, Patterson D, Koenen MA, Greenfield TK (2008). Alcohol content variation of bar and restaurant drinks in Northern California. Alcoholism: Clinical and Experimental Research, 32(9):1623-1629. Read on PubMed →
- Kwok A, Dordevic AL, Paton G, Page MJ, Truby H (2019). Effect of alcohol consumption on food energy intake: a systematic review and meta-analysis. British Journal of Nutrition, 121(5):481-495. Read on PubMed →
- Suter PM, Jequier E, Schutz Y (1994). Effect of ethanol on energy expenditure. American Journal of Physiology, 266(4 Pt 2):R1204-R1212. Read on PubMed →
- Suter PM, Schutz Y, Jequier E (1992). The effect of ethanol on fat storage in healthy subjects. New England Journal of Medicine, 326(15):983-987. Read on PubMed →
- Schutz Y (2000). Role of substrate utilization and thermogenesis on body-weight control with particular reference to alcohol. Proceedings of the Nutrition Society, 59(4):511-517. Read on PubMed →
- Sayon-Orea C, Martinez-Gonzalez MA, Bes-Rastrollo M (2011). Alcohol consumption and body weight: a systematic review. Nutrition Reviews, 69(8):419-431. Read on PubMed →
- National Heart, Lung, and Blood Institute, NIH (1998). Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. NIH Publication 98-4083. Read the guidelines →
The formulas and limits behind MyPlate's own numbers, with their citations, are on the health sources page.