The right way to measure your dog for a harness.

About forty per cent of harness returns happen because the dog was measured wrong. Not because the harness was bad. Not because the sizing chart was misleading. Because someone — usually the owner, sometimes the manufacturer — skipped a measurement or took it while the dog was in the wrong position.

This is avoidable. A harness fitting takes three measurements, two minutes, and a soft tape. The rules are specific enough to be learned once and then applied to any brand you'll ever buy from. Here they are.

Fair warning: if you've been measuring only the chest girth, you have been measuring one of three numbers you needed, and you've been getting lucky when the harness fits.

Measurement one: chest girth

This is the measurement most sizing charts start with, and the one most owners already know. Wrap a soft tape around the widest part of the ribcage, just behind the front legs. The tape should lie flat against the fur, snug but not compressed — you should be able to slide two fingers under it comfortably.

Two mistakes are common here. First, measuring too far back, over the soft belly rather than the ribcage. The ribcage is harder, broader, and stable; the belly narrows as the dog moves and changes size with every breath. Take the measurement at the ribs.

Second, pulling the tape too tight to get a "true" number. A snug-but-breathing measurement is the number that will fit. A compressed number will put you in a harness size down, and the harness will rub.

Measurement two: neck base

Most owners skip this. Most returns come from skipping this.

The neck measurement is taken at the base of the neck — where the neck meets the shoulders — not up under the jaw where a collar sits. This is the point where the front straps of most harnesses cross, and a harness that fits the chest but is wrong here will ride up, chafe, or slip.

Wrap the tape around the widest part of the neck just above the shoulders. Same rule: two fingers of give. For deep-chested breeds (greyhounds, whippets, dobermans) the neck base can be wider than the head, and for these dogs some harness styles simply won't work regardless of chest size — you'll need a step-in or a bespoke cut.

Measurement three: back length

Nose-to-tail is not useful. Back length — the measurement you actually need — is taken from the base of the neck (where you just measured) to the base of the tail, following the spine.

This determines whether the harness will sit correctly or extend past the ribs onto the soft belly. A harness that's too long for the back will interfere with the back legs on long strides and shift around during movement. Too short and it will sit too high, putting pressure on the shoulders.

Most sizing charts don't ask for back length, but the good ones list it as a reference figure for each size. If yours doesn't, email the maker and ask. A brand that can't tell you the back-length of each size probably doesn't know.

The position rule nobody mentions

All three measurements must be taken while the dog is standing, weight evenly on all four legs, in a relaxed posture. Not sitting. Not lying. Not while being held. Not right after exercise when the ribs are expanded.

Dogs measured while sitting come out two to three centimetres smaller on chest girth than the same dog standing. Dogs measured lying down come out four to five centimetres smaller. Those differences are the gap between a size Medium and a size Small.

Wait until the dog is calm, stand them on a flat surface, distribute weight on all four legs, and measure. If the dog moves, start again. Two minutes of patience here saves a return.

Reading sizing charts honestly

Sizing charts are presented as clean tables. In practice they are negotiations between the manufacturer and the pattern of returns they've accepted as normal. Three rules for reading them:

When your measurement is between two sizes, size up. A harness slightly too large can be adjusted tighter with the straps. A harness slightly too small cannot be adjusted looser. Most harness straps have four to six centimetres of adjustability in either direction.

Trust the measurement range, not the breed table. Charts often list breeds ("Labrador — size L; Beagle — size M"), which is roughly correct on average but wrong for a specific dog. Your labrador might be lean or barrel-chested. Go by the numbers.

Deep-chested breeds need deep-chested brands. A greyhound measured as "size M" on a standard chart will often need size L in chest and size S in neck on the same harness. Brands that make harnesses specifically for sighthounds account for this. General-purpose brands usually don't.

Three measurements, two minutes, one calm dog. It is not complicated. It is just specific.