Milk or water: which one stays in you, and what milk adds

Swapping water for milk or juice at a meal raised the total energy people ate by about 14.9 percent (range 10.9 to 23.9) in a systematic review of single-meal tests, and its authors call that evidence much less conclusive, with milk and juice counted together. The same 2010 review says randomized trials are still needed before anyone can claim that choosing water over milk helps control weight. Where milk does measurably more is holding on to fluid. In a randomized crossover trial in 16 college students who cycled in the heat until they lost about 2 percent of body weight, skimmed lactose-free milk was retained at about 69 percent over 3 hours against about 40 percent for water. In another small crossover trial at rest, a liter of low-fat milk stayed in the body longer than water in 12 young men and did not in 11 older men. A cup of whole milk (244 g) brings about 149 calories and 300 mg of calcium by the US dietary survey, a figure scaled from 100 g values, and a cup of tap water brings 0 calories.

How much is actually in it

• Per 100 g, whole milk carries about 123 mg calcium (range 116 to 130 mg in 11 samples) and nonfat milk about 132 mg, while plain tap water carries no energy. — USDA Foundation Foods analyzed samples of fluid milk (n = 11-12 samples per milk type) Evidence E source
  Calcium: whole 123.0 mg (116-130, n=11); 1% 126.0 mg (111-143, n=12); 2% 126.0 mg (111-134, n=12); nonfat 132.0 mg (118-153, n=12) per 3.5 oz (100 g)

What a real portion looks like

• In the US dietary survey, 1 cup of whole milk (244 g) supplies about 149 calories and 300 mg calcium, while 1 cup of tap water (240 g) supplies 0 calories. — USDA FNDDS portion weights, Milk, whole (code 11111000) and Water, tap (code 94000100) Evidence E source
  Whole milk 1 cup 244 g: 148.8 calories, 300.1 mg calcium; tap water 1 cup 240 g: 0 calories

What your body absorbs

• In a randomized crossover trial in 16 active students, skimmed lactose-free milk drunk after exercise in the heat was retained at about 69% versus about 40% for water over 3 hours. — 16 college students, cycling in the heat to lose about 2% body mass, then drinking 150% of the loss (n = 16) Evidence B source
  Fluid retention about 69% (milk) vs about 40% (water), p < .001; net fluid balance -0.57 lb (-0.26 kg) vs -1.5 lb (-0.67 kg) at 3 h
• In a randomized crossover trial in 15 men after exercise, cow's milk was retained at 40% versus 16.6% for a sports drink after 4 hours, though milk caused more bloating and fullness. — 15 young men who lost 2% body mass by intermittent cycling, then drank 150% of the loss over 1 h (n = 15) Evidence B source
  Beverage retained: cow's milk 40.0% ± 24.9%, soy milk 46.9% ± 19.9%, Powerade 16.6% ± 16.5%

What cooking and storage change

• Porridge made with whole milk has about 112 calories, 4.9 g protein and 131 mg calcium per 3.5 oz (100 g), against 47 calories, 1.4 g protein and 6 mg calcium when made with water. — CoFID 2021 recipes: porridge made with water (11-1107) and made with whole milk (11-1108) Evidence E source
  Per 100 g, milk vs water: energy 112 vs 47 calories; protein 4.9 vs 1.4 g; fat 4.7 vs 1.0 g; total sugars 4.8 g vs trace; calcium 131 vs 6 mg

What it does to your health

• A systematic review of clinical trials estimated that drinking milk or juice instead of water with a meal raised total energy intake by about 14.9%, though the evidence was weak. — people in clinical trials comparing a beverage with water at a test meal Evidence C source
  Total energy intake +14.9% (range 10.9 to 23.9%) when water was replaced with milk or juice

Safety, limits and interactions

• In the same trial, drinking about 2.1 quarts (2 liters) of skimmed lactose-free milk within 90 minutes caused only mild gut symptoms, no different from the same volume of water. — 16 college students after exercise in the heat, about 2.1 quarts (2 L) of each drink over 90 min (n = 16) Evidence B source
  No difference in gastrointestinal symptoms between milk, water and sports drink (p > .05)
• In a Cochrane review of 16 trials in 1467 young children with acute diarrhea, lactose-free products instead of lactose-containing milk shortened diarrhea by about 18 hours. — children under five with acute diarrhea, mostly inpatients in high- or middle-income countries (n = 1467) Evidence A source
  MD -17.77 hours (95% CI -25.32 to -10.21), low quality evidence

What people believe that the data does not support

• A systematic review found no benefit from drinking 8 glasses of water a day for people without kidney stones, and extra fluid can worsen overactive bladder. — studies from 1972 to 2017 on water intake and disease risk Evidence C source
  No evidence of benefit outside nephrolithiasis; excess fluid can exacerbate overactive bladder symptoms

Who this works differently for

• In a randomized crossover trial of 1.1 quarts (1 L) of different drinks, low-fat milk was retained longer than water in young men, but this advantage was not seen in older men. — 12 young and 11 older men, euhydrated, drinking 1.1 quarts (1 L) of water, fruit juice, sports drink or low-fat milk (n = 23) Evidence B source
  Milk retained longer than other beverages in young but not older men; sodium balance negative after water at 3 h, neutral after milk
• For people who avoid dairy, a randomized crossover trial in 10 men found a sweetened soya drink kept the body as hydrated as skim milk over 3 hours. — 10 healthy euhydrated fasted males, 34 fl oz (1000 mL) water delivered as skim milk or sweetened soya beverage over 30 min (n = 10) Evidence B source
  Total urine mass 986 ± 254 g (milk) vs 950 ± 248 g (soya), P = 0.435
• Across national surveys, plain water made up to 58% of beverage intake in children versus up to 80% in adults, and milk intake was highest in children. — 65 nationwide or city-representative surveys of healthy children, adolescents and adults, 2000 to 2013 Evidence C source
  Water up to 58% (children), 75% (adolescents), 80% (adults) of total beverage intake

What is still unknown

• Beyond single-meal tests showing about 15% more energy eaten when milk or juice replaces water, whether choosing water over milk helps control weight remains unproven, as randomized trials are still lacking. — clinical, epidemiologic and intervention studies comparing water with other beverages Evidence C source
  Evidence gap: no confirming randomized trials

The bottom line

Milk also changes the food you cook in it. By the UK food tables, porridge made with whole milk has 112 calories, 4.9 g protein and 131 mg calcium per 3.5 oz (100 g), against 47 calories, 1.4 g protein and 6 mg calcium when made with water, both figures calculated from recipes. Skimming does not cost calcium. USDA samples put whole milk at 123 mg per 3.5 oz (100 g) (116 to 130 across 11 samples) and nonfat milk at 132 mg (118 to 153 across 12). The rule of eight glasses of water a day was checked. A systematic review of studies from 1972 to 2017 found no benefit for people without kidney stones, and extra fluid can make an overactive bladder worse. That review mixed study designs and pooled no numbers. Against a sports drink after exercise, cow's milk was retained at 40 percent and the sports drink at 16.6 percent over 4 hours in 15 young men, with a wide spread between individuals and more bloating and fullness on milk. If you avoid dairy, a sweetened soya drink matched skim milk in a crossover trial in 10 men, with 950 g against 35 oz (986 g) of urine over 3 hours, and that trial had no water arm. Across 65 national and city surveys, plain water made up at most 58 percent of what children drank and at most 80 percent for adults, and milk intake was highest in children. On safety, 16 students drank about 2.1 quarts (2 liters) of skimmed lactose-free milk within 90 minutes and reported only mild gut symptoms, no different from water. Because that milk was lactose-free, the result says nothing about ordinary milk for people who digest lactose poorly. For children under five with acute diarrhea, a Cochrane review of 16 trials in 1467 children, mostly hospital inpatients, found lactose-free products shortened the illness by about 18 hours (95% CI 10 to 25 hours) compared with lactose-containing milk, on low quality evidence. We found no card on cow's milk allergy, on pregnancy or on medicines for this pair.

Sources

1. mwt-daily-r5-11 · reference dataset
   Milk, whole, 3.25% milkfat: Calcium 123.0 mg, range 116.0-130.0, n=11 · Milk, lowfat 1%: 126.0 mg, range 111.0-143.0, n=12 · Milk, reduced fat 2%: 126.0 mg, range 111.0-134.0, n=12 · Milk, nonfat: 132.0 mg, range 118.0-153.0, n=12
   USDA FoodData Central Foundation Foods, release 2026-04-30 · checked 2026-10-03
2. mwt-daily-r5-12 · reference dataset
   Milk, whole (fdc 2705385): 1 cup 244 g, Energy 148.8 kcal, Calcium 300.1 mg | Water, tap (fdc 2710707): 1 cup 240 g, Energy 0.0 kcal
   USDA FNDDS, codes 11111000 and 94000100 · checked 2026-10-03
3. mwt-daily-r5-01 · randomized crossover trial
   SLM showed higher fluid retention (∼69%) versus W (∼40%; p < .001); SD (∼56%) was not different from SLM or W (p > .05). Net fluid balance was higher for SLM (-0.26 kg) and SD (-0.42 kg) than W (-0.67 kg) after 3 hr (p < .001), resulting from a significantly lower diuresis with SLM.
   Int J Sport Nutr Exerc Metab, 2024 · checked 2026-10-03
4. mwt-daily-r5-03 · randomized crossover trial
   Upon completion of trials, the percentage of beverage retained was Sustagen Sport 65.1% ± 14.7%, soy milk 46.9% ± 19.9%, cow's milk 40.0% ± 24.9%, and Powerade 16.6% ± 16.5%. ... Subjective ratings of bloating and fullness were higher during all milk trials compared with Powerade whereas ratings of overall thirst were not different between beverages.
   Appl Physiol Nutr Metab, 2014 · checked 2026-10-03
5. mwt-daily-r5-13 · reference dataset
   CoFID 11-1107 Porridge, made with water: Energy 47 kcal · Protein 1.4 g · Fat 1.0 g · Total sugars Tr · Calcium 6 mg | CoFID 11-1108 Porridge, made with whole milk: Energy 112 kcal · Protein 4.9 g · Fat 4.7 g · Total sugars 4.8 g · Calcium 131 mg
   CoFID 2021, Public Health England (OGL v3.0), codes 11-1107 and 11-1108 · checked 2026-10-03
6. mwt-daily-r5-06 · systematic review
   Much less conclusive evidence was found in studies replacing water with milk and juice, with estimated increases in TEI of 14.9% (range, 10.9 to 23.9%).
   Nutr Rev, 2010 · checked 2026-10-03
7. mwt-daily-r5-02 · randomized crossover trial
   Reported GI disturbances were mild and showed no difference among drinks (p > .05) despite ingestion of W (1,992 ± 425 ml), SD (1,999 ± 429 ml), and SLM (1,993 ± 426 ml) in 90 min.
   Int J Sport Nutr Exerc Metab, 2024 · checked 2026-10-03
8. mwt-daily-r5-10 · meta-analysis of RCTs
   Compared to lactose-containing milk, milk products, or foodstuffs, lactose-free products may reduce the duration of diarrhoea by an average of about 18 hours (MD -17.77, 95% CI -25.32 to -10.21, 16 trials, 1467 participants, low quality evidence).
   Cochrane Database Syst Rev, 2013 · checked 2026-10-03
9. mwt-daily-r5-08 · systematic review
   The available reviewed literature suggests no benefit to drinking 8 glasses of water per day in patients without nephrolithiasis. Also, excess fluid intake can exacerbate symptoms of overactive bladder.
   J Urol, 2018 · checked 2026-10-03
10. mwt-daily-r5-04 · randomized crossover trial
    At 3 h Na+ balance was negative after ingesting water and fruit juice, but neutral after sport drink and milk. ... Milk was retained longer than other beverages in Young, but not in Older, despite similar net electrolyte balance responses.
    Eur J Appl Physiol, 2023 · checked 2026-10-03
11. mwt-daily-r5-05 · randomized crossover trial
    Total urine mass (MILK, 986 ± 254 g; SOYA, 950 ± 248 g; P = 0.435) and urine specific gravity (P = 0.156) did not differ between trials.
    Eur J Appl Physiol, 2024 · checked 2026-10-03
12. mwt-daily-r5-09 · systematic review
    Plain water contributed up to 58%, 75% and 80% of the total beverage intake in children, adolescents and adults, respectively. Milk consumption was higher among children; consumption of soft drinks was higher among adolescents; and the consumption of tea, coffee and alcoholic beverages was higher among adults.
    J Hum Nutr Diet, 2015 · checked 2026-10-03
13. mwt-daily-r5-07 · systematic review
    The available literature for these comparisons is sparse and somewhat inconclusive. ... A need for randomized-controlled trials to confirm this role exists.
    Nutr Rev, 2010 · checked 2026-10-03