How we classify body shape

The method uses bust, waist, hips and shoulders. Most calculators ask for three and drop the shoulder, which is why they cannot separate an inverted triangle from an hourglass: both have a waist that is clearly narrower than the ends, and the only thing distinguishing them is whether the top of the frame outmeasures the hip.

None of the four is used on its own. Every rule below is a comparison between two of your own numbers, so the unit cancels and the classification is the same in inches or centimeters. This is also why a shape is not a size: a ratio does not know how large the measurements are, only how they relate.

Five tests run top to bottom, and the first one to fire is the answer. The order is the part most calculators get wrong.

Tests 1 and 2 ask which end of the frame is wider. Tests 3 and 4 ask how defined the waist is. Dominance is checked first on purpose: a figure measuring 36-33-40 has hips four inches past the bust and a waist that is not especially drawn. Run the waist test first and it returns rectangle, which is the wrong advice for a body that is visibly wider at the hip. Run dominance first and it returns pear, which is right.

Rectangle is last because it is the residual, not a positive finding. It is what you get when neither end dominates and the waist is neither notably defined nor notably full — the middle of the distribution, and the most common single answer.

There is no published standard for these thresholds. No sizing body, medical authority or academic group defines where a pear becomes a rectangle, and anyone who implies otherwise is inventing an authority that does not exist. Body shape is a styling convention, not a taxonomy.

The three numbers used here — 1.05 for dominance, 0.85 for a full waist and 0.75 for a defined one — were fitted to the ratios that the widely used calculators already agree on, so a figure classified here gets the same answer it would get elsewhere. The 5% dominance margin exists because measurement error is real: a padded bra or a tape held half an inch off level moves a reading by more than an inch, and a threshold with no tolerance would flip shapes on noise.

If enough real measurements suggest the cut-offs sit in the wrong place, they are the two numbers to move, and this page will say so when they change.

Every tool on this site runs the five rules above in the same order — what differs is only how the comparisons are read. The calculator reads them from a tape measure. The quiz reads them by eye in a mirror. The photo analyzer reads them from your silhouette: it finds 33 pose landmarks in one full-body photo, measures the silhouette's width at the bust, waist, high-hip and hip lines, and converts each width to its tape-measure equivalent before running the identical rules. The conversion factors come from ANSUR II, the US Army's public anthropometric survey (1,986 women measured in both circumference and width): a silhouette under-reads the bust, whose tape measurement lives partly in depth, and over-reads the shoulders, which are wide but flat. The photo is analyzed on your device and never uploaded.

The instruments differ in what they can see. The quiz and the photo analyzer can both detect the upper-hip shelf, so they reach all six shapes where the calculator reaches five. The photo analyzer reads width but not depth, so fullness carried forward reads narrower than a tape would find it — and rather than guess, it rejects photos whose pose or clothing would corrupt the reading and says what to change. When instruments disagree, trust the tape.

The two studies this site leans on, so the figures can be checked rather than taken on faith:

Who publishes this and under what rules is on the about page.