String, paper, printout or app: every ring sizing method ranked

Every method for finding a ring size trades accuracy against convenience, and the range is wider than most people assume. The gap between the best and worst options on this page is more than a full ring size, which on a plain band is the difference between comfortable and unwearable.

Here is the honest ranking, including the cases where reaching for an app is the wrong answer.

Which ring sizing method is most accurate?

The pattern is worth noticing before the detail. Everything in the top half measures a rigid object against a verified scale. Everything in the bottom half measures a flexible thing against a scale nobody checked.

Why is a jeweller’s mandrel the reference standard?

A mandrel is a tapered steel cone marked with sizes, usually paired with a set of steel rings to try on. It is the physical standard that every other method on this page is approximating, it costs nothing to use, and the whole thing takes about two minutes.

Its one weakness is timing rather than technique. It measures you at whatever moment you happen to walk in, which includes the cold January afternoon when your fingers are at their smallest of the year. A jeweller who knows their trade will notice and say something. Not all of them do.

Go in person when the ring is expensive, when it cannot be resized, or when it is a full eternity band where getting it wrong means remaking rather than adjusting.

How accurate is measuring a ring you already own?

This is the best method available at home, and it is close enough to the mandrel that the difference rarely matters.

You place a ring on a screen that knows its own pixel density and match a circle to the band’s inner edge. What makes it work is that you have eliminated both of the big error sources at once. The thing you are measuring is rigid, so it does not change size while you look at it, and the scale you are measuring against is exact, because the app set it from the device’s published pixel density before drawing anything.

The catch is that you need a ring that fits the correct finger, and the app has to genuinely be calibrated. An app that assumes one generic screen density can be several sizes out without giving any sign of it, which is covered in how screen calibration works.

Are digital calipers worth buying?

If you already own a pair, use them. Open the small inside jaws within the band until they touch both walls and read the display. It is precise, objective and takes about ten seconds, with no interpretation required.

The obvious limitation is that most people do not own calipers and are not going to buy them for one ring. They also only work on an existing ring, so they solve nothing if you are measuring a bare finger. That said, a basic digital pair costs less than a single resize, which makes them worth considering if you buy jewellery with any regularity.

How accurate is measuring a finger with an app?

About half a size, and the limit is the finger rather than the tool.

Everything about the scale is exact. The variable is that a finger is not the same size twice: it changes with the time of day, the temperature, what you ate, whether you have exercised, and several things you cannot control. That sets a floor on the accuracy of any finger measurement regardless of how good the instrument is.

You improve it the same way regardless of method. Measure in the evening, at room temperature, with warm hands, three times, and take the largest of the three.

When does a printed sizer actually work?

When you check it, which almost nobody does.

Print at one hundred percent scale, then hold a bank card against the reference rectangle on the page. It should measure 85.60 mm wide, because that is the ISO/IEC 7810 standard every payment card is manufactured to. If it matches, your printout is true and the circles on it are accurate. If it does not, reprint with scaling switched off.

The problem is not that printing is a bad idea. It is that a printed sizer’s accuracy is unknown until you verify it, while a screen that reads its own pixel density starts from a known scale and can be checked in two seconds with the same bank card.

What is wrong with tracing and string?

Tracing works if you hold the pencil vertically, and fails quietly if you do not. A pencil held at forty-five degrees traces a circle about a millimetre wider than the ring it is going round, which is a size and a half of error produced entirely by wrist angle. Nothing about the tracing looks wrong afterwards.

The paper strip method is honest and free. You wrap, mark, and measure with a ruler. Its limits are how hard you pull and how well anyone reads a millimetre off a household ruler, which puts it around a full size in either direction.

String is the one to avoid. It is recommended more widely than any other method and it is the worst on the list, because string is a textile and textiles stretch under tension. Two millimetres of stretch is most of a ring size, and you cannot feel two millimetres.

When is an app the wrong answer?

Three situations, and it is worth being direct about them.

Go to a jeweller for full eternity bands, tension settings, and non-resizable metals like tungsten or ceramic. When a mistake means remaking the ring rather than adjusting it, a quarter of a size of uncertainty is not good enough, and the trip is cheap insurance.

Go to a jeweller if your knuckle is much larger than the base of your finger. A two-size gap is not a measurement problem, it is a fitting problem, and it needs sizing beads, a spring insert or a hinged shank rather than a better number. The big knuckles guide covers the options.

Go to a jeweller for arthritic or swollen joints, where a hinged shank is usually the right answer and no amount of measuring will substitute for it.

What is the best combination?

Measure a ring you already own on a calibrated screen, then measure the finger with the on-screen gauge as a cross-check.

If the two land within half a size of each other, order with confidence. If they disagree by more than that, something is wrong, and it is usually that the reference ring came from a different finger than you assumed. The point of the second measurement is not extra precision. It is catching a mistake before you spend money rather than afterwards.

The same logic applies to a printed sizer, incidentally. If you have already printed one and checked it against a bank card, use it as your second opinion rather than throwing it away. Two independent methods that agree are worth far more than one method you have repeated three times, because repeating a method three times mostly confirms that the method is consistent, not that it is right.

Where the two disagree and you cannot work out why, that is the point to spend two minutes in a jewellery shop. You are no longer asking them to do the job for you. You are asking them to break a tie, which takes a jeweller about twenty seconds and settles the question for good.