A hydrometer is one of the simplest instruments in homebrewing — a weighted glass float and a printed scale — and that simplicity is exactly why so few brewers get formally taught how to read one correctly. Most people watch someone else do it once, copy the general idea, and carry small, compounding reading habits forward for years without ever checking whether they're right. The instrument itself is trustworthy. The technique around it is where the error creeps in.
Why this matters more than it seems to
Every number this site's calculators produce starts from a gravity reading, and none of the downstream math can correct for a reading that was wrong at the source. A 1.050 OG to 1.012 FG pair computes to 4.99% ABV by the simple formula, 5.08% by the advanced one, with 76.0% apparent attenuation. Check it here. That result is only as good as the two readings feeding it — a technique error worth a couple of thousandths on either end shifts every number in that result, not just the one you were focused on.
Mistake one: reading the wrong part of the meniscus
Liquid in a narrow tube curves at the surface — this is the meniscus, and where you read it matters. The correct read is at the bottom of the curve, at eye level, looking straight across rather than down at an angle. Reading at the top of the curve, or looking down at the sample from above, both introduce a small but real bias, and the direction and size of that bias depends on exactly how the reading was mistaken — which is precisely why it's such an easy, invisible error to carry forward for years without ever catching it. Get your eye level with the liquid surface, read where the flat plane of liquid (not the curved edge climbing the glass) crosses the scale, every time.
Mistake two: not accounting for temperature
Hydrometers are calibrated to read accurately at one specific reference temperature, commonly 60°F or 68°F — a sample read at any other temperature needs correcting before the number means anything as a true gravity value. This is covered in full, with a complete correction table, on the hydrometer temperature correction chart, but the short version is worth internalising here: a sample pulled hot straight out of an actively boiling or freshly-cooled kettle can be off by several points if read and recorded without correcting or cooling first. The fix is simple — cool the sample close to your hydrometer's marked calibration temperature before reading whenever you can, and correct explicitly when you can't.
Mistake three: reading a sample that still has dissolved CO2 in it
A sample pulled from an actively fermenting batch carries dissolved CO2 that's still actively escaping as tiny bubbles cling to and lift the hydrometer, producing an unstable, artificially inconsistent reading that can shift visibly while you're watching it. The fix is a gentle swirl of the sample in its testing tube (never a vigorous shake, which risks oxidation) to encourage the CO2 to release before you take the reading, or simply drawing the sample a few minutes ahead of when you actually plan to read it and letting it settle. A hydrometer that won't sit still, drifting up and down slightly as you watch, is telling you it's still degassing — wait it out rather than recording whatever number it happens to show at that moment.
A fourth, less obvious one: never having checked your own hydrometer
Cheap hydrometers vary unit to unit, and a hydrometer that's carrying a small, consistent manufacturing offset will quietly bias every single reading you ever take with it in the same direction, forever, unless you catch it. The check is simple and takes two minutes: read plain, cool distilled or filtered water. A correctly calibrated hydrometer should read almost exactly 1.000 at its marked calibration temperature (some genuine, tiny variation is normal and fine). If yours consistently reads 1.002 or 0.998 on plain water instead, note that offset and apply it as a correction to every future reading, or replace the instrument if the offset is large enough to matter for your purposes.
Why a hydrometer works the way it does
Worth understanding briefly, because it explains why the technique matters as much as it does: a hydrometer floats at a depth determined by the density of the liquid around it — denser liquid (more dissolved sugar) supports more of the float's weight higher up, so the instrument rides higher and the scale reads a higher number. It's a purely physical measurement with no moving parts and nothing to calibrate electronically, which is exactly why it's so reliable and exactly why it's so sensitive to the handful of physical conditions covered above — temperature changes the liquid's density independent of sugar content, dissolved gas changes the effective buoyancy transiently, and a misread meniscus is simply human error layered on top of an otherwise perfectly reliable measurement.
Hydrometer vs. refractometer, briefly
A refractometer measures gravity a completely different way — by how light bends passing through a few drops of sample — and it has real advantages a hydrometer doesn't (a tiny sample size, and a fast reading with no need to fill and clean a full testing tube). But a refractometer introduces its own, larger correction requirement once fermentation starts, because dissolved alcohol confuses its optics in a way sugar concentration alone doesn't — see refractometer correction, explained for how large that effect actually is. Many experienced brewers use a refractometer for quick progress checks during active fermentation, specifically because it only needs a couple of drops rather than a full sample pulled off the batch, but fall back to a hydrometer for the final, decisive FG reading, since a hydrometer needs no alcohol correction at all and is the more directly trustworthy instrument for the number that actually matters most.
Building the habit of recording, not just reading
A single accurate reading is only half the value — recording it, with the date and the sample temperature if you didn't cool it to calibration temperature first, turns a one-off number into a trend you can actually use. Two readings a few days apart, both taken with the same careful technique, tell you whether fermentation is still progressing or has genuinely finished — see why is my final gravity stuck at 1.020 for exactly this decision. A single reading, however carefully taken, only ever tells you where things stood at one moment; the habit of reading the same careful way every time is what makes a series of readings mean something.
Putting it together
None of these four fixes takes more than a few extra seconds once they're habits rather than something you have to consciously remember. Read at eye level, at the bottom of the meniscus. Correct for temperature, or cool the sample first. Let dissolved CO2 clear before recording a number. And know your own instrument's baseline before trusting it on a real batch. The payoff isn't a dramatically different-looking number most of the time — it's that when a reading genuinely does look surprising (a stuck fermentation, an unexpectedly high or low FG), you can trust that the surprise is real and worth investigating, rather than wondering whether it's just technique noise. See why is my final gravity stuck at 1.020 for exactly that situation, and apparent vs. real attenuation for the next layer of what a clean, trustworthy pair of readings actually tells you once you're confident they're accurate.