What total body water means
Total body water is all the water in the body, inside cells and outside them, and it is the largest single component of body mass. The commonly quoted adult reference spans are 50 to 65 percent of body weight for men and 45 to 60 percent for women. The difference between those spans is not really about sex; it is about composition. Muscle is about 75 percent water while adipose tissue is 10 to 20 percent, so anyone with a higher body fat share has a lower water share, and average body fat differs between men and women.
The Watson equation was built by measuring total body water directly with deuterium dilution — the subject drinks a known quantity of labelled water, it distributes through the body's water, and the resulting concentration reveals the volume it distributed into — and then fitting a regression to height, weight, age and sex. The male form kept an age term, the female form did not, because age was not statistically significant in the female data.
Read the error before the decimal place
The reported standard error for the Watson equation is around 3.8 litres for men. Against a typical value in the low forties, that is close to ten percent. The output shows a figure to one decimal place because that is how these numbers are conventionally presented, but the decimal is decoration — the real statement is "somewhere in a four-litre-wide band, probably".
The same caution applies down the page. Deurenberg's body fat estimate carries about four percentage points of individual error, so a result of 22 percent is consistent with anything from 18 to 26. Boer and Hume are shown together specifically so their disagreement is visible, because that gap is a lower bound on the uncertainty rather than a curiosity.
The 73 percent cross-check
Lean tissue holds a fairly stable water fraction, conventionally taken as 73 percent. That constant links two of the estimates on this page: divide the Watson body water figure by 0.73 and you get an independent route to lean body mass, which can be compared with what Boer returns.
When the two land within a few percent of each other, the inputs are behaving like the populations these equations were fitted on. When they separate, something is off — most often lean mass far from average in either direction. The check is worth running precisely because it can fail. What it cannot do is confirm correctness: the equations were derived from similar adult samples in a similar era, so they are capable of being wrong in the same direction at the same time, and passing the check does not rule that out.
Why a bioimpedance scale disagrees
A bioimpedance device passes a small alternating current through the body and measures the resistance, which is lower through water-rich tissue than through fat. That resistance then goes into a regression of its own to produce water and composition figures. So it is also an estimate from a formula — the difference is that one of its inputs is a physical measurement made on you, which lets it track individual variation that a height-and-weight equation cannot.
The trade is sensitivity to hydration state. The same person measured before and after a meal, before and after training, in the morning versus the evening, or at different points of a menstrual cycle, can shift by two to three percentage points of body fat on the same device. Formula estimates are perfectly stable across the day, which sounds like an advantage until you notice it also means they cannot detect anything that actually changed. The two approaches fail in opposite ways, which is a reasonable argument for looking at both and trusting neither absolutely.
This is not a hydration recommendation
Body water is how much water is in the body now. Daily fluid requirement is a separate question with separate determinants — climate, activity, diet, kidney function, medication. A low estimated water share does not mean drink more; it mostly means a higher body fat share, and drinking extra water changes the ratio not at all, because the kidneys excrete the surplus within hours. The lever on body water percentage is lean mass, and it moves over months.
Questions people ask
My scale gives a different body water percentage. Which is right?
Neither is a measurement, so the question does not have a clean answer. The scale uses a real electrical measurement taken on your body, which lets it capture individual differences a height-and-weight formula cannot, and that is a genuine advantage. Against that, it is sensitive to hydration state and moves by two to three points across a single day for the same person. The formula is stable but blind to everything except height, weight, age and sex. They are different kinds of estimate with different failure modes rather than a right answer and a wrong one.
Should I drink more water if the percentage looks low?
No, that inference does not follow. Body water percentage is set overwhelmingly by body composition, because muscle carries about 75 percent water and fat tissue only 10 to 20 percent. Someone with a higher body fat share will read lower regardless of how much they drink, and additional intake is excreted within hours rather than raising the ratio. Fluid needs are a separate question driven by climate, activity, diet and medical factors. If you have reason to think your fluid balance is genuinely off — persistent swelling, unusual thirst, sudden weight change — that is a medical question, not a calculator one.
Why does the female Watson equation not use age?
Because in the original 1980 dataset the age term was not statistically significant for the female subjects, so it was dropped from the final fitted equation. That is a statement about that data, not a claim that body water is age-independent in women. The most likely reading is that height and weight already accounted for most of the age-related variation in that particular sample. Differences of this kind between formulas usually trace back to the composition of the sample they were fitted on rather than to biology.
I train seriously and the body fat estimate looks far too high. Why?
That is the well-documented failure mode of the Deurenberg equation. Its only body input is BMI, so weight from muscle and weight from fat enter the calculation identically, and someone carrying substantial muscle reads high. In that situation the lean mass equations and the water-derived cross-check on this page are more informative than the fat percentage line, and a tape-based estimate that responds to shape — the body fat calculator — will usually read considerably lower. When several estimates spread widely, that spread is the finding.
When is the best time to measure?
For a device measurement, first thing in the morning after using the bathroom and before eating or drinking gives the most repeatable conditions. Straight after a meal, after training, or after a bath all shift fluid distribution enough to move the reading. Formula-based estimates like the ones on this page are unaffected by timing, but that also means they cannot register any real change that happened during the day. Either way, a series of readings taken under identical conditions tells you more than any single value.