Jackson & Pollock 3-site vs 7-site skinfold: which to choose

By PounceFit Team · Updated on 2026-10-06

Technical content for health and fitness professionals. It reproduces formulas and protocols published in the scientific literature, with the sources at the end of the page. It is not a recommendation and does not replace the judgment of the professional who runs the assessment.

Between the Jackson & Pollock 3-site and 7-site skinfold protocols, accuracy is similar: both versions come from the same studies, and in the original samples they estimated body density with similar accuracy (Jackson and Pollock, 1978; Jackson, Pollock and Ward, 1980). Use the 3-site version when time is short or the client prefers to expose less of the torso, and the 7-site version when you want to dilute the effect of one poorly measured fold and cover more body regions. Consistency matters more than the choice: the same protocol, the same assessor and the same caliper at every reassessment.

Both calculations are in the 3-site and 7-site calculators. In PounceFit, body assessment software for personal trainers, both versions are available, and the 7-site version is the default.

Where the two versions come from

Jackson and Pollock (1978) measured skinfolds, circumferences and body density (by hydrostatic weighing) in 308 men aged 18 to 61 and validated the equations in a separate sample of 95 men. Using the quadratic form of the skinfold sum, plus age, correlations exceeded 0.90, with a standard error of about 0.0073 g/ml. In the validation sample, the error was close to 0.0077 g/ml.

Jackson, Pollock and Ward (1980) repeated the design with 249 women aged 18 to 55. The 3-, 4- and 7-site equations had correlations of 0.842 to 0.867 and standard errors of 3.6 to 3.8 percentage points of body fat. In the validation sample of 82 women, the error was 3.7 to 4.0 points. The authors advised caution with women over 40.

In other words: in the original data, going from 3 to 7 sites did not reduce the error in any meaningful way. A well-chosen 3-site sum already carries much of the information in the 7-site sum.

The formulas in full

S is the sum of the protocol's skinfolds, in millimeters. Age is in years.

7-site (chest, midaxillary, triceps, subscapular, abdomen, suprailiac and thigh):

• Men: density = 1.112 − 0.00043499 × S + 0.00000055 × S² − 0.00028826 × age
• Women: density = 1.097 − 0.00046971 × S + 0.00000056 × S² − 0.00012828 × age

3-site, men (chest, abdomen and thigh):

• Density = 1.10938 − 0.0008267 × S + 0.0000016 × S² − 0.0002574 × age

3-site, women (triceps, suprailiac and thigh):

• Density = 1.0994921 − 0.0009929 × S + 0.0000023 × S² − 0.0001392 × age

In both versions, density is converted to percent body fat with the Siri (1961) equation:

Body fat (%) = (4.95 ÷ density − 4.50) × 100

PounceFit uses exactly these coefficients and the Siri conversion in both versions.

Side-by-side comparison

Worked example: both calculations on the same client

Man, 30 years old, 80 kg (176 lb). Skinfolds (mean of measurements, in mm): chest 10, midaxillary 12, triceps 11, subscapular 15, abdomen 22, suprailiac 16 and thigh 14. This is the same client as in the 7-site skinfold article.

With 7 sites:

1. S = 10 + 12 + 11 + 15 + 22 + 16 + 14 = 100 mm.
2. Density = 1.112 − 0.043499 + 0.0055 − 0.0086478 = 1.0653532.
3. Body fat = (4.95 ÷ 1.0653532 − 4.50) × 100 = 14.6%.

With 3 sites (chest, abdomen and thigh):

1. S = 10 + 22 + 14 = 46 mm; S² = 2,116.
2. Density = 1.10938 − 0.0008267 × 46 + 0.0000016 × 2,116 − 0.0002574 × 30 = 1.10938 − 0.0380282 + 0.0033856 − 0.007722 = 1.0670154.
3. Body fat = (4.95 ÷ 1.0670154 − 4.50) × 100 = 13.9%.

From the same pinches, the two versions differ by 0.7 percentage points. Neither is "wrong": each equation has its own error against hydrostatic weighing. The problem shows up when the client is assessed with one version and reassessed with the other. A drop from 14.6% to 13.9% would look like a change in body composition, when it is only a change of equation.

In women the gap can go the other way. A 25-year-old woman with triceps 18, suprailiac 14, thigh 25, chest 8, midaxillary 10, subscapular 14 and abdomen 20 mm comes out at 21.8% with 7 sites (S = 109 mm) and 22.8% with 3 sites (S = 57 mm). The full 3-site calculation is in the Jackson & Pollock 3-site calculator.

When to choose the 3-site version

• Short on time, or many assessments in a row. It takes 3 sites instead of 7.
• Client who prefers not to expose the torso. In women, the 3-site version skips chest, midaxillary, subscapular and abdomen.
• Assessor still in training. Fewer sites to find and mark. But each error weighs more in the result, so train your technique before you trust the numbers.

When to choose the 7-site version

• When technique varies. A 2 mm error at one site changes a 46 mm sum by more than 4%; in a 100 mm sum, by 2%.
• When you want to see distribution. The 7 sites cover torso, arm and thigh. Looking site by site helps you read where the change happens between assessments.
• When your practice already uses 7 sites. Switching protocols breaks your clients' history.

Limits of both versions

• Source population. General-population adults in the United States, within the age ranges above. Outside them, and in groups with very different body composition, the error tends to grow (Heyward and Wagner, 2004).
• The equations are not interchangeable. In 2,458 active adults, different formulas applied to the same skinfolds gave mean body fat values far apart from each other, and the authors recommend always using the same formula when results will be compared (Vaquero-Cristóbal et al., 2020). That study did not include Jackson & Pollock, but the logic holds for any change of equation.
• Measurement sites. Follow the sites described by the ACSM (2021) and always measure on the right side. The suprailiac site in particular moves depending on the standard you adopt.
• Tracking. The skinfold sum in millimeters tracks change without depending on any equation (Heyward and Wagner, 2004).