Photo vs DEXA, calipers and BIA: which body fat method is best?

By ScanBodyFat. Published October 26, 2026.

Short answer

No single method is perfect: DEXA, hydrostatic weighing and the Bod Pod are lab or clinic methods, calipers depend on the person using them, BIA scales depend on hydration, and the tape formula and photo AI are convenient estimates with a few points of error. For a quick, free starting point a photo estimate is reasonable; for a clinical number, use a clinic.

The methods side by side

Figures below come from different studies on different groups (athletes, Marines, general adults), so they are not directly comparable. Where a number is not available from a source we link, the table uses qualitative wording.

Comparison of body fat measurement methods
Method	Access and cost	Error and caveats	Pros and cons
DEXA	Clinic or imaging center, scheduled and paid. Described as relatively low expense compared with MRI or CT.	Technical error of about 1.9% for total fat mass in one review; results shift with food, hydration and prior exercise.	Widely used reference and gives limb detail. Uses a very small radiation dose, and needs standard pre-scan conditions.
Hydrostatic (underwater) weighing	Uncommon and costly equipment, mostly in labs.	Assumes tissue densities; residual lung volume is an error source.	Historically a reference method. Time consuming and can be uncomfortable.
Bod Pod (air displacement)	Expensive; mainly found in research and some university or clinic settings.	High reliability in one review (CV 1.7%); read at least 1.28% higher than hydrostatic weighing; sensitive to clothing, hair and moisture.	Quick and non-invasive. Diverges from DXA at the extremes of BMI.
Skinfold calipers	Inexpensive and quick; needs a trained person.	Depends on site and technician; over 100 conversion equations exist, and one athlete's result ranged from 4% to 8% depending on the equation.	Good for tracking change when the same person measures the same sites. Better reported as a sum of millimeters.
BIA scales	Low cost and widely available.	Large variability between devices; affected by hydration and meal timing. In one study, consumer scales averaged about 4.5 to 5.9 points of error and professional BIA about 3.1 to 4.7, against DXA.	Convenient. May help track change under identical conditions, less so as an absolute number.
Tape / Navy formula	Free; needs a tape measure and a helper for best results.	In a Marine sample, average bias vs DXA was about minus 2.5 points for men and plus 1 to 2 points for women, with wide spread.	Simple and free. Authors call it a field method, not research grade.
Photo AI	Free or low cost, from home, about a minute.	About 2.2 points in one study of a purpose-built system under controlled photos; about 4.1 in an earlier research method. ScanBodyFat states roughly 3 to 5 points.	Fast and no equipment. Depends on photo quality and has more error than clinic methods.

What the sources say, method by method

DEXA is described as the current reference method for body composition and as relatively low in expense compared with MRI or CT (Marra et al.). A sports review reports a technical error of about 1.9% for total fat mass and notes that a meal can raise measured fat mass by as much as 2.6% (narrative review).
Hydrostatic weighing and the Bod Pod are lab methods. The same review calls underwater weighing uncommon, costly and potentially uncomfortable, and says the Bod Pod is expensive and mainly used in research.
Skinfold calipers are inexpensive and quick, but the review stresses that results depend on the measurement site and the technician, and recommends tracking the sum of skinfolds over time.
BIA accuracy is reduced when proper equations and standardized protocols are not used (Marra et al.). In the 2022 photo study, consumer and professional BIA devices had mean absolute errors of roughly 3.1 to 5.9 percentage points against DXA, compared with 2.16 for that study's photo system (Majmudar et al.).
The tape (Navy) formula was evaluated in a convenience sample of 430 men and 179 women in the US Marine Corps; the authors describe it as not suitable for applications requiring quantitative body composition (Potter et al.).
Photo AI results depend on the system. One research method reported a mean absolute error of about 4.1% in 323 people (Affuso et al.); another reported 2.16 points in 134 adults (Majmudar et al.). Neither is a study of ScanBodyFat.

Which should you use?

For a medical or clinical need, ask a doctor and use a clinic method.
For tracking change over months, pick one method and keep conditions identical.
For a quick, free starting point with no equipment, a photo estimate is reasonable, as long as you remember the margin. See how accurate photo estimates are and what typical ranges mean.

A number from any method is one data point, not a health verdict, and lower is not automatically better.

Frequently asked questions

Which body fat measurement method is most accurate?

A published review states there is no gold-standard body composition method in free-living individuals. DEXA is considered by many to be the criterion standard, and hydrostatic weighing was historically the reference, but both depend on conditions and have their own error.

Are BIA scales accurate for body fat?

Results vary a lot. Reviews describe large variability between devices and sensitivity to hydration, and in one study consumer scales had mean absolute errors of roughly 4.5 to 5.9 percentage points against DXA. They may be more useful for tracking change under consistent conditions than for an absolute number.

Is a photo estimate as accurate as DEXA?

No. DEXA is a clinical measurement. A photo estimate is a convenient range that carries roughly 3 to 5 percentage points of error on ScanBodyFat, and about 2 points in one published study of a different purpose-built system under controlled conditions.

Is the tape measure (Navy) formula reliable?

It is described by its evaluators as a field method for classification, not for research-grade measurement. A study of US Marines found it underestimated body fat for men and overestimated it for women compared with DXA, on average.

Want a number for your own photo?

Free, no signup. Upload one photo and a few details, get an estimated range.

Get your free estimate

Sources

More guides

Estimates are for fitness purposes only, not a medical measurement or diagnosis. If thinking about body measurements feels stressful, you can skip them. For support with eating or body image concerns, see the National Eating Disorders Association.