Pure water at 20 degrees Celsius and standard atmospheric pressure has a density of approximately 0.998 g/cm³, or about 998.2 kg/m³. That number sits just below the convenient round figure of 1.000 g/cm³ that many people remember from school, and the gap between the two tells a surprisingly rich story about how water behaves, how the metric system was designed, and why small decimal places matter in science and industry.
Why the Number Is Not Exactly 1.000
The reason most people expect water’s density to be a clean 1.000 g/cm³ traces back to how the metric system was originally constructed. In the late eighteenth century, the kilogram was defined to approximate the mass of one liter of pure water at its temperature of maximum density. That maximum occurs near 4°C, not 20°C. At 3.98°C, water reaches its peak density of about 0.99997 g/cm³, tantalizingly close to but not quite 1.00000. The original kilogram artifact was crafted to match this value as closely as the technology of the time allowed.1Metrologia. A brief history of the unit of mass: continuity of successive definitions of the kilogram By 20°C, water has warmed well past that peak and expanded slightly, which is why its density drops to about 0.998.
So the “water equals one” shorthand works as a rough estimate, but it was only ever meant to be exact at one specific temperature and under ideal conditions. At room temperature, the discrepancy is small enough that it hardly matters for everyday purposes like cooking or filling a fish tank. It matters enormously, though, in laboratory calibrations, pharmaceutical manufacturing, and precision engineering.
How Temperature Shifts Water’s Density
Water’s density does not change at a constant rate as temperature rises. Near freezing, the relationship is unusual: warming water from 0°C actually makes it denser, not less dense, until it hits that peak around 4°C. Above 4°C, water behaves the way you would expect any liquid to behave, expanding and becoming less dense as it gets warmer. By 20°C, the density has dropped to roughly 0.998 g/cm³. By 50°C it falls to about 0.988. At the boiling point under standard pressure, it is down to approximately 0.958.
The strange behavior below 4°C is one of water’s most consequential quirks. It means ice floats, and it means the coldest water in a lake sinks only to a point before the very coldest water rises back toward the surface and freezes there. This density anomaly is driven by the way water molecules link to one another through hydrogen bonds. As water cools toward freezing, its molecules begin arranging themselves into a more open, hexagonal lattice structure, the same structure that defines ice. That open arrangement takes up more space per molecule, so the density drops even as the temperature continues to fall. Research on the hydrogen-bond network suggests that the shifting balance between tightly packed and loosely bonded molecular clusters accounts for the density peak without requiring any exotic structural phases.2PubMed. Correlation of structural order, anomalous density, and hydrogen bonding network of liquid water
At temperatures above 4°C, including 20°C, the picture is simpler. More thermal energy means more molecular motion, which pushes molecules apart on average and reduces density steadily. The rate of decrease is gentle at first and accelerates as the water approaches boiling. For anyone working with water in a temperature-controlled laboratory at 20°C, the key takeaway is that even a degree or two of temperature drift produces a measurable density change, on the order of a few tenths of a kilogram per cubic meter per degree Celsius in that range.
What Dissolved Substances Do to the Value
The 0.998 g/cm³ figure assumes pure water with nothing dissolved in it. Real water almost always contains something extra, and those additions change the density.
The most dramatic everyday example is seawater. Ocean water at a salinity of about 35 parts per thousand and a temperature of 20°C has a density close to 1.025 g/cm³, roughly 2.5 percent higher than pure water at the same temperature. The relationship between salinity, temperature, and density has been mapped in detail for oceanography; early reference tables date to 1902, and more precise measurements over the decades have refined those values. One study covering salinities from 9 to 41 parts per thousand across temperatures from 0 to 25°C found that the older reference tables were slightly low, by an average of about 0.006 in sigma-t units, a small but meaningful correction for deep-sea research and climate modeling.3Deep Sea Research and Oceanographic Abstracts. The specific gravity/salinity/temperature relationship in natural sea water
Even plain tap water contains dissolved minerals like calcium and magnesium, which bump its density slightly above the pure-water value. Freshwater lakes and rivers vary depending on their mineral load, though the differences are usually in the fourth or fifth decimal place and irrelevant outside of precise measurement work.
Less obvious is the effect of dissolved air. Water exposed to the atmosphere absorbs small quantities of nitrogen, oxygen, and other gases. This dissolved air lowers the density of water very slightly compared to fully degassed (air-free) water. The effect has been studied from 0°C to 50°C by calculating how the dissolved gas components shift the water’s overall mass-to-volume ratio.4International Journal of Thermophysics. Effect of Dissolved Air on the Density and Refractive Index of Water For most purposes the dissolved-air correction is negligible, amounting to a few parts per million. But in high-precision calibration labs, where instruments are validated to five or six significant figures, it matters enough that some standards specify whether the reference density applies to air-saturated or air-free water.
Heavy Water and Isotopic Variation
Not all “pure” water is identical at the molecular level. Most of the hydrogen in ordinary water is protium, with a single proton and no neutrons. A small fraction is deuterium, which carries one neutron. Water made entirely with deuterium, called heavy water or D₂O, is about 11 percent denser than ordinary water at the same temperature, with a density near 1.105 g/cm³ at 20°C. Natural water contains roughly 0.015 percent deuterium, so its density is shifted upward by a tiny amount compared to a hypothetical sample of perfectly isotope-free water.
This matters in fields like nuclear physics and isotope geochemistry, where heavy water is used as a neutron moderator in certain reactor designs and as a tracer in hydrological studies. The reference equation of state for heavy water has been refined separately from ordinary water because D₂O’s thermodynamic properties differ enough that you cannot simply scale the normal water equations.5Journal of Physical and Chemical Reference Data. A Reference Equation of State for Heavy Water For someone asking about “the density of water at 20°C,” the quoted 0.998 g/cm³ assumes ordinary water with a natural isotopic mix, which is the standard in virtually all practical contexts.
When Precision Actually Matters
For everyday life, rounding water’s density at 20°C to 1.00 g/cm³ introduces an error of about 0.2 percent. If you are estimating whether a container will overflow or figuring out how much a water bottle weighs, that margin is meaningless. But several industries and scientific disciplines depend on knowing the density to many more decimal places.
Calibration of volumetric equipment is probably the most widespread application. Pipettes, graduated cylinders, flow meters, and proving tanks are all validated by weighing a known volume of water and working backward to the volume using the water’s density at the measured temperature. A published guide on selecting water-property formulas for these calibrations emphasizes that the same density equations serve both the highest-level metrological laboratories and routine industrial flow measurement.6Metrologia. The selection of water property formulae for volume and flow calibration Getting the density wrong by even a few parts in ten thousand propagates into every volume measurement downstream. Pharmaceutical companies, for instance, need to dispense precise volumes of liquid reagents; if their pipettes are calibrated against the wrong density value, every dose measured with those pipettes carries a systematic error.
Brewing, winemaking, and distilling also rely on water-density knowledge, though for a different reason. In these industries, density measurements of the final product (compared to pure water at a reference temperature) are used to estimate alcohol content, sugar content, or both. A hydrometer floating in a wine sample reads higher or lower depending on how the dissolved sugars and alcohol shift the liquid’s density relative to water. The reference point for that comparison is typically water at 20°C, so any error in the reference density feeds directly into the alcohol-by-volume calculation on the label.
Environmental monitoring is another area where water density comes into play. Oceanographers tracking water masses and ocean currents need density profiles down to the fifth decimal place because tiny density differences between adjacent water layers drive the large-scale circulation of the ocean. Limnologists studying lakes use the same principles on a smaller scale: the density difference between warm surface water and cold deep water determines how and when a lake mixes, which in turn controls oxygen distribution and nutrient cycling.
Pressure’s Role
The standard density value assumes one atmosphere of pressure, the average pressure at sea level. Water is often described as “incompressible,” but that is only approximately true. Under the pressures found deep in the ocean or in industrial hydraulic systems, water’s density increases measurably. At the bottom of the Mariana Trench, where pressures exceed 1,000 atmospheres, water is roughly 5 percent denser than at the surface. For any application at or near sea level, the pressure effect is negligible, but deep-ocean and high-pressure industrial work needs to account for it.
Even altitude matters at the margins. At high elevations, atmospheric pressure is lower, which slightly reduces the boiling point and subtly changes the equilibrium of dissolved gases. These effects nudge the density by amounts that are irrelevant for all but the most sensitive calibrations.
Supercooled Water and Extreme Conditions
Water does not always freeze at 0°C. In the absence of nucleation sites like dust particles or container imperfections, liquid water can be cooled well below freezing without solidifying, a state called supercooling. Supercooled water is denser than ice at the same temperature, but its density changes in interesting ways as it gets colder. Research on the volumetric properties of supercooled water shows that the difference in molar volume between liquid water and ice shrinks as the temperature drops, eventually approaching zero near a critical temperature of about 228 K (roughly −45°C).7PubMed Central. Activity of Supercooled Water on the Ice Curve and Other Thermodynamic Properties of Liquid Water up to the Boiling Point at Standard Pressure At that point, the liquid and solid phases become nearly indistinguishable in volume, even though they remain structurally different.
This is far removed from the 20°C reference condition, but it illustrates just how strange water is as a substance. Most liquids contract smoothly and predictably with cooling. Water’s density curve bends, peaks, and does unexpected things at both ends of its liquid range, all because of the geometry of hydrogen bonds.
Common Misconceptions About Water Density
A few misunderstandings come up regularly when people look into this topic. The first is treating 1.000 g/cm³ as the universal density of water at any temperature. As described above, 1.000 only approximates the value near 4°C, and even there it falls slightly short of a perfect 1.00000. At 20°C the gap is small but real.
A second misconception is thinking that “pure water” in a laboratory sense and tap water have the same density. They are close, but municipal water supplies contain dissolved minerals, chlorine or chloramine, and dissolved gases that collectively push the density slightly higher or lower than the textbook pure-water value. For most purposes this does not matter, but it is worth knowing if you ever try to verify a density measurement at home with a kitchen scale and a measuring cup: your tap water is not the same substance the reference tables describe.
A third is assuming that hot water is always less dense than cold water. That is true above 4°C, but below 4°C the relationship reverses. Someone pouring near-freezing water into a container of slightly warmer water at, say, 6°C might expect the cold water to sink. It will, but only because it is still above 4°C. If the cold water were at 2°C, it would actually float on top of the 4°C layer, because it is less dense despite being colder.
Measuring Water Density at Home
If you wanted to check the 0.998 g/cm³ figure yourself, the simplest approach is to weigh a known volume of water at a known temperature. Fill a volumetric flask (or any container whose volume you trust) with water that has been sitting at room temperature long enough to equilibrate, weigh the full container, subtract the container’s weight, and divide mass by volume. A good kitchen scale reads to one gram, which limits you to a very rough check, but a lab-grade balance reading to a tenth of a gram and a carefully calibrated 1,000 mL flask can get you within a few tenths of a percent of the reference value.
The tricky parts are controlling temperature and accounting for dissolved air. Water fresh from the tap has been pressurized and often holds more dissolved gas than water that has been left sitting open. Warming and cooling the sample even slightly during measurement introduces error. Professional labs use temperature-controlled baths and carefully degassed water for exactly these reasons, and they apply published density formulas that correct for measured temperature down to the hundredth of a degree.
For anyone who just needs a quick number for a school project or rough calculation, 0.998 g/cm³ at 20°C or the even simpler 1.00 g/cm³ will serve perfectly well. The extra decimal places exist for the people whose work depends on them, and the science behind those decimals reveals water to be a far stranger liquid than its familiarity might suggest.