Water's density is about 997 kilograms per cubic meter at 4 °C, which works out to roughly 0.997 grams per cubic centimeter or 62.4 pounds per cubic foot. That maximum value holds only near 4 °C at standard atmospheric pressure. Warm it to boiling and it drops to about 958 kg/m³; freeze it and the solid floats at roughly 917 kg/m³. Everything else in this article hangs on those reference conditions, because "the density of water" without a stated temperature is an incomplete answer.
Why 4 °C Is the Reference Temperature for Water Density
Most liquids contract steadily as they cool, so their density climbs all the way to the freezing point. Water breaks that pattern between 0 °C and 4 °C. As liquid water cools from room temperature toward 4 °C it contracts like any other liquid, but below 4 °C it starts expanding again. That is why the maximum sits at 4 °C rather than at the freezing point.
The cause is hydrogen bonding. Water molecules in the liquid state continually form and break short-lived bonds that pull them closer together. As the liquid cools, those clusters grow and pack more tightly — until around 4 °C, where the geometry of the hydrogen-bond network starts arranging molecules into the open hexagonal pattern that eventually becomes ice. Once that open structure begins to dominate, further cooling adds volume instead of removing it.
This is not a curiosity. It's why a pond freezes from the top down. Water chilled to 4 °C near the surface sinks because it's denser than the warmer water below, and the mixing that follows distributes oxygen and nutrients through the water column. When the surface finally reaches 0 °C and freezes, that ice layer insulates the liquid beneath it, which stays close to 4 °C and keeps aquatic life alive through winter.
The 4 °C point is a reference, not a universal truth. It shifts slightly with pressure and with dissolved substances, which is why the standard definition pairs 4 °C with one atmosphere. For everyday calculations — aquarium volumes, recipe conversions, pipe sizing — those shifts are negligible. For calibration work or oceanographic modeling, they are not.
The Full Density Table From 0 °C to 100 °C
The figures below are for pure water at one atmosphere, based on values published by the USGS Water Science School and standard engineering tables. Liquid water stays within about 2.5% of its maximum density across the entire range from freezing to boiling — a tighter spread than most people expect.
| Temperature | Density (kg/m³) | Density (g/cm³) | Density (lb/ft³) |
|---|---|---|---|
| 0 °C (liquid, just above freezing) | 999.84 | 0.99984 | 62.42 |
| 4 °C | 999.97 | 0.99997 | 62.43 |
| 10 °C | 999.70 | 0.99970 | 62.41 |
| 20 °C | 998.21 | 0.99821 | 62.30 |
| 25 °C | 997.05 | 0.99705 | 62.24 |
| 40 °C | 992.22 | 0.99222 | 61.94 |
| 60 °C | 983.20 | 0.98320 | 61.38 |
| 80 °C | 971.80 | 0.97180 | 60.67 |
| 100 °C | 958.40 | 0.95840 | 59.83 |
Read the 0 °C row carefully. Liquid water right at the freezing point is already slightly less dense than water at 4 °C — the anomaly starts before any ice forms. The 999.84 kg/m³ value applies to supercooled liquid water held just above 0 °C at one atmosphere; once ice crystals appear, you're in a different material with a different density.
The practical spread is small but real. A 200-liter hot water tank at 80 °C holds about 194 kg of water, while the same tank filled cold at 10 °C holds nearly 200 kg. That 6-kilogram difference is why hot water systems need expansion tanks — the same mass of water occupies more volume as it heats, and the pressure has to go somewhere.
Density vs. Specific Gravity vs. Weight Density: Sorting the Terms
Density is mass per unit volume, full stop. Specific gravity is a ratio — the density of a substance divided by the density of water at a stated reference temperature. Because the two numbers are so close for water itself, specific gravity of water is essentially 1.000 at 4 °C, which is precisely why the scale was built that way.
Weight density, sometimes called specific weight, is weight per unit volume rather than mass per unit volume. In metric practice it's expressed in newtons per cubic meter; in US customary units it's pounds per cubic foot, which is the 62.4 lb/ft³ figure most American engineering references quote. Confusing weight density with mass density is the single most common error in unit conversion, and it shows up constantly in plumbing and brewing calculations.
One more term worth pinning down: relative density. It's used interchangeably with specific gravity in most contexts, though some standards reserve "relative density" for comparisons where both substances are at the same temperature. For water quality work, the distinction rarely changes a number by more than 0.001.
The misuse to watch for is treating "density of water = 1 g/cm³" as exact. It's a rounding that's fine for a kitchen or a classroom and wrong for a laboratory. The true value at 4 °C is 0.99997 g/cm³, and the difference — about 3 parts in 100,000 — matters when you're calibrating a volumetric flask or computing the mass of a precise water sample. Use 1.000 for estimation, 0.99997 for calibration.
Water vs. Ice vs. Steam: The Same Substance at Three Densities
Ice at 0 °C comes in at roughly 917 kg/m³, about 8.3% less dense than liquid water at the same temperature. That gap is enormous by the standards of phase changes, and it's entirely due to the hexagonal crystal lattice that hydrogen bonds lock into as water freezes. The open structure traps empty space between molecules.
Steam is a different story. At 100 °C and one atmosphere, saturated water vapor has a density of about 0.598 kg/m³ — roughly 1,600 times less dense than liquid water. That ratio is what makes steam such an effective volume-filling medium in heating systems and why a small amount of liquid water flashing to steam can pressurize a large vessel.
| Phase | Conditions | Density (kg/m³) | Relative to liquid at 4 °C |
|---|---|---|---|
| Ice | 0 °C, 1 atm | ~917 | 91.7% |
| Liquid water | 4 °C, 1 atm | 999.97 | 100% |
| Liquid water | 100 °C, 1 atm | 958.4 | 95.8% |
| Saturated steam | 100 °C, 1 atm | ~0.598 | 0.06% |
The ice comparison has a practical edge case that catches people out: ice formed under pressure. Glacial ice at the bottom of a thick ice sheet is compressed to densities above 920 kg/m³, and the deepest ice in Antarctica can exceed 920 kg/m³ as air bubbles are squeezed out. That's still less dense than liquid water, so the ice sheet floats — but the density is not the textbook 917 figure. If you're working with glacier ice or sea ice, use a measured value, not the standard number.
A second edge case: ice formed from seawater. Sea ice starts as fresh-water crystals that exclude salt, so its bulk density depends on how much brine is trapped in the pore spaces. First-year sea ice typically runs 900–920 kg/m³, while older multi-year ice can be denser. The number you need depends on the ice type, not just the temperature.
How Temperature and Pressure Actually Move the Number
Temperature is the dominant variable across the range humans encounter. Going from 4 °C to 100 °C cuts density by about 4.2%, and going from 4 °C down to 0 °C cuts it by only 0.013%. The curve is asymmetric, steep at the hot end and nearly flat at the cold end, which is why cooling water past 4 °C gains you almost nothing in density terms.
Pressure matters far less than most people assume. Water is close to incompressible, so squeezing it from one atmosphere to 100 atmospheres — roughly the pressure at 1,000 meters depth in the ocean — increases density by only about 0.5%. At the bottom of the Mariana Trench, around 1,086 atmospheres, water is about 4.5% denser than at the surface. That's a large pressure by any human standard and a modest density change by any physical standard.
Salinity is the variable that catches people who work with real water rather than pure water. Seawater at 3.5% salinity and 4 °C has a density near 1,028 kg/m³, about 2.8% higher than fresh water. That difference drives thermohaline circulation, the deep-ocean conveyor that moves heat around the planet, and it's the reason a ship's draft changes when it moves from a river to the open sea.
For a worked example, consider a swimming pool 10 meters long, 4 meters wide, and 2 meters deep — 80 cubic meters of water. At 25 °C (997.05 kg/m³) it holds 79,764 kg. Fill it with 10 °C water (999.70 kg/m³) and it holds 79,976 kg, a difference of 212 kg, roughly the mass of three adults. Same pool, same volume, different mass — because density changed by 0.27%.
Reading Density Off a Hydrometer in the Field
A hydrometer measures specific gravity directly by floating at a depth determined by the liquid's density, and it's the standard field tool for brewing, aquarium keeping, and battery testing. The reading you get is specific gravity relative to water at a calibration temperature printed on the instrument — often 20 °C or 60 °F.
Here's the procedure for a correct reading, in the order that prevents the common errors:
1: Bring the sample to the hydrometer's calibration temperature, or note the actual temperature and apply the correction chart that ships with the instrument. 2: Fill a clean, tall cylinder to within 2–3 cm of the rim, avoiding bubbles; let it stand until foam disperses. 3: Lower the hydrometer gently into the liquid and spin it slightly to release any clinging bubbles. 4: Read the scale at the bottom of the meniscus, eye level with the surface. 5: Record the value and the temperature together, then convert to density by multiplying specific gravity by the density of water at the calibration temperature.
The failure mode that ruins most readings: a sample that hasn't degassed. Dissolved carbon dioxide in a fermenting beer sample will lift the hydrometer and read low, sometimes by 0.002–0.003 specific gravity points — enough to throw off an alcohol calculation by half a percent. Let the sample go flat before measuring.
A second failure case is the hydrometer that's been dropped. A cracked or chipped bulb changes the displacement and shifts every reading. Check the instrument against distilled water at the calibration temperature before trusting it; it should read 1.000, and if it doesn't, the offset tells you how far off it is.
What Density Does to Buoyancy and Everyday Life
An object floats when its average density is lower than the water it's in, and the fraction submerged equals the ratio of the two densities. A block of oak at roughly 700 kg/m³ floats with 70% of its volume underwater. A block of ice at 917 kg/m³ floats with about 92% submerged, which is why icebergs hide most of their mass below the surface.
This is a calculation you can do at the dock. A 4-meter aluminum boat with a hull volume of 3 cubic meters displaces 3,000 liters of water, and at freshwater density near 1,000 kg/m³ that's about 3,000 kg of buoyant force. Subtract the boat's own mass and you get the payload before it swamps. Change to seawater and the same hull floats about 2.8% more mass, which is why boats sit noticeably higher in the ocean than in a lake.
The density difference between hot and cold water also runs the thermosiphon, the passive circulation loop in solar water heaters and some wood-stove systems. Warm water rises through the collector, cools in the tank, and sinks back — no pump required. The driving force is the density gap between the hot and cold legs of the loop, which grows with temperature difference. At a 40 °C spread, the density difference is about 26 kg/m³, enough to drive a slow but reliable flow.
In the kitchen, density explains why oil floats on water (most cooking oils run 910–930 kg/m³) and why a dense syrup sinks through a lighter one. It also explains the failure case every baker hits once: substituting by volume instead of weight. A cup of water weighs about 237 grams at room temperature, while a cup of flour weighs 120–125 grams. The density difference is why "one cup" means different masses in different ingredients, and why a kitchen scale beats a measuring cup every time.
The Numbers That Trip People Up: Common Misconceptions
The most persistent misconception is that water is densest at 0 °C because that's its freezing point. It isn't. Densest liquid water sits at 4 °C, and both warming and cooling from there reduce density. The anomaly is the whole reason lakes don't freeze solid from the bottom up.
A second error is treating 1 kg/L as exact. One liter of water at 4 °C has a mass of 0.99997 kg, not 1.000 kg. The 1 kg/L convention is a rounding built into the original definition of the kilogram and the liter, and it's off by about 3 parts per 100,000. That's fine for almost everything and wrong for metrology.
A third is assuming density and viscosity move together. They don't. Honey is far denser than water and far more viscous, but hot water is slightly less dense and much less viscous than cold water. Viscosity is about resistance to flow; density is about mass per volume. Mixing them up leads to bad pump sizing and worse coffee.
The fourth is the unit trap. US customary density of water is often quoted as 62.4 lb/ft³, but that's weight density, not mass density, and the two diverge the moment you leave Earth's gravity. On the Moon, water's mass density is unchanged at about 997 kg/m³, but its weight density drops to roughly one-sixth of the terrestrial figure. Use mass density for physics, weight density for engineering in a gravity field, and don't swap them.
FAQ
Is the density of water exactly 1 g/cm³?
No. The precise value at 4 °C is 0.99997 g/cm³, and at 20 °C it's 0.99821 g/cm³. The "1 g/cm³" figure is a convenient rounding that's accurate to about 0.3% across normal room temperatures. It's fine for cooking and rough engineering; use the tabulated value when you're calibrating glassware or computing masses to more than three significant figures.
Why does ice float if it's made of the same water?
Ice floats because freezing locks water molecules into an open hexagonal lattice, which leaves more empty space per molecule than the liquid state does. The result is ice at 0 °C with a density near 917 kg/m³ floating on liquid water at 999.84 kg/m³. About 92% of a floating ice cube sits below the waterline, which is why icebergs are a hazard well beyond their visible profile.
Does saltwater have the same density as freshwater?
No. Seawater at typical ocean salinity of 3.5% and 4 °C runs about 1,028 kg/m³, roughly 2.8% denser than fresh water at the same temperature. That difference drives deep-ocean circulation and changes how vessels float between river and sea. A ship moving from fresh to salt water rises in the water by about 2.8% of its draft.
What is the density of water in pounds per gallon?
A US gallon of water at 4 °C weighs about 8.34 pounds, which corresponds to a density of roughly 62.4 lb/ft³. At 100 °C the same gallon weighs about 8.0 pounds because the water has expanded by about 4%. The 8.34 lb/gal figure is the one to use for plumbing and irrigation estimates at ordinary temperatures.
Can water be denser than 1,000 kg/m³?
Yes, under pressure or with dissolved solids. At the bottom of the Mariana Trench, around 1,086 atmospheres, water is about 4.5% denser than at the surface, putting it near 1,042 kg/m³ for cold seawater. Heavy brine solutions can exceed 1,200 kg/m³ at ordinary pressure. Pure fresh water at one atmosphere never exceeds 999.97 kg/m³.
What happens to water density between 0 °C and 4 °C?
Density rises as water warms from 0 °C to 4 °C, going from 999.84 to 999.97 kg/m³. That's a tiny change — about 0.013% — but it reverses the normal cooling-contracts pattern. Below 4 °C, cooling causes expansion; above 4 °C, cooling causes contraction. The reversal is driven by hydrogen-bond clustering.
Does water density change with altitude?
Barely. At 3,000 meters elevation, atmospheric pressure drops to about 70% of sea level, and pure water's density falls by roughly 0.01–0.02% as a result. That's within the rounding of most tables. The bigger altitude effect is on boiling point, which drops to about 90 °C at 3,000 meters, shifting the hot end of the density curve.
Why do hot water systems need expansion tanks?
Because water expands as it heats, and the density drop from 10 °C to 80 °C means the same mass occupies about 2.8% more volume. In a closed system that extra volume has nowhere to go and drives pressure up. An expansion tank gives it a place to push, absorbing the change before it stresses pipes, valves, or the water heater itself.