Skinfold Body Fat: the 3-Site and 7-Site Protocols

Caliper testing estimates body fat by pinching skin and the fat directly under it at standardized points, reading each thickness in millimeters, and running the sum through a regression equation. Done by a trained tester it is one of the better field methods. Done by yourself in a bathroom mirror it produces a confident number that means nothing.

That is why this page explains the method instead of offering a form to fill in.

The 3-site and 7-site protocols

The generalized equations by Jackson and Pollock remain the most widely used. They exist in a short form and a long form, and the sites are not interchangeable between sexes.

Protocol	Men	Women
3-site	Chest, abdomen, thigh	Triceps, thigh, suprailium
7-site	Chest, midaxillary, triceps, subscapular, abdomen, suprailium, thigh	Same seven sites

The short version exists for speed and the long version for accuracy, and mixing sites between them invalidates both — the coefficients were fitted to those specific sums, not to skinfolds in general.

From millimeters to a percentage

The equations do not output body fat directly. They predict body density from the sum of folds, adjusted for age, and a second equation converts density to a percentage. That second step is almost always the Siri equation:

BF% = [(4.95 ÷ body density) − 4.5] × 100

Two conversions, two sets of assumptions. The density model assumes fixed densities for fat and fat-free tissue, which vary with age, hydration and bone mineral content — so even a flawless pinch inherits error from the arithmetic that follows it.

How much error the calipers carry

The 2022 systematic review in Current Obesity Reports found strong evidence that the Jackson-Pollock equations are valid for estimating total body fat mass in adults, which is real support for the method.

The magnitude is the caveat. A 2023 validation against a three-compartment criterion model reported standard errors of 1.9 to 4.6 kg in females and 3.3 to 7.6 kg in males, and the authors concluded that skinfold-derived measures do not yield accurate fat and fat-free mass values, particularly when used to track change. StatPearls is blunter still, describing skinfold testing as having high variability and limited clinical use so far.

Read that as a boundary rather than a dismissal: calipers place you in a band, they do not detect the two pounds of fat you lost last month.

Getting a measurement worth having

• One tester, one caliper, throughout. Switching either resets your history.
• Always the same side of the body, conventionally the right.
• Pinch skin and fat only, never muscle, and read at a consistent count after releasing the jaws.
• Three readings per site, median rather than mean if one is an outlier.
• Not straight after training, when the skin is engorged and folds compress differently.

When there is no trained tester

Two self-serve methods stay reproducible without a second person. The body fat percentage calculator needs only weight, height and age. The Navy body fat calculator uses a tape at the neck and waist and handles muscular builds better than any BMI-derived estimate. Neither matches good caliper work for accuracy, and both beat bad caliper work by a wide margin.

For competitors, where the decision justifies the cost, a repeated DXA scan on the same machine is the standard — see body fat percentage for athletes for what those scans return sport by sport.

This content is informational and does not replace evaluation by a health professional.