Are Smart Scales Accurate for Body Fat?

The short answer: smart scales are accurate for body weight and inaccurate for body fat. When researchers compared consumer smart scales with DEXA, the scales measured weight to within a median of 0 to 0.3 kg, but underestimated fat mass by a median of 2.2 to 4.4 kg. They’re also very repeatable, though, so a smart scale used under identical conditions can still show you a useful trend.

What the studies found

Three consumer smart scales vs. DEXA

A 2021 study in JMIR mHealth and uHealth weighed hospital patients on a consumer smart scale immediately after a DEXA scan. It compared three scales (Téfal Body Partner, Terraillon DietPack and Withings Body Cardio), with 48 to 53 people per scale.

Measure	Median error vs. DEXA
Body weight	0 to 0.3 kg
Fat mass	−2.2 to −4.4 kg (underestimated)

The authors’ conclusion was blunt: smart scales are not accurate for body composition. For someone weighing 80 kg, a 3 kg error in fat mass shifts body fat percentage by almost 4 points.

Fifteen impedance devices vs. a lab reference

A 2023 study in the British Journal of Nutrition tested fifteen bioelectrical impedance devices, fourteen consumer-grade and one research-grade, on 73 adults. Each was compared with a laboratory four-compartment model, which combines several reference methods.

• Repeatability was excellent. Precision error ranged from 0.0% to 0.49%.
• Single-reading accuracy varied widely. Average error ranged from −3.5 to +11.7 percentage points depending on the device.
• Tracking change was better. For people re-tested 12–16 weeks later, error in the change was much smaller. Octapolar models (hand grips plus foot pads) and some foot-to-foot scales did best.

Why smart scales get body fat wrong

Smart scales use bioelectrical impedance. A small current travels from one foot to the other, and the scale measures how much the body resists it. Lean tissue holds a lot of water and conducts well; fat holds little and resists more. The scale runs that resistance, plus the height, age and sex you entered, through a prediction equation.

Two problems follow:

1. The equation is a population average. If your build differs from the people the equation was built on, the estimate is off, consistently.
2. Body water moves constantly. Hydration, a salty dinner, a high-carb day, alcohol, a workout, even cold feet all change the current’s path without changing your fat. That’s also why your weight changes before your body does.

A foot-to-foot scale also mostly “sees” your legs and lower body, then extrapolates to the rest of you.

How to get a useful trend out of a smart scale

• Same time, same state. First thing in the morning, after the bathroom, before food, drink, coffee or training.
• Same scale, same surface. Hard floor, never carpet, and never compare across brands.
• Weekly averages, not days. Average five to seven readings and compare week to week.
• Watch direction, not level. If your average drifts from 24% to 22% over two months under identical conditions, that direction is more trustworthy than either number.
• Cross-check occasionally with a different method: a waist measurement, the tape-measure or skinfold calculator, or comparable progress photos.

Where a smart scale fits among the other methods

A smart scale is the most convenient home method and among the least accurate for an absolute number. DEXA is the most accurate and the least practical to repeat. Skinfolds depend on the measurer: one small Brazilian study found skinfolds agreed better with underwater weighing than four impedance devices did. Photo-based estimates trade equipment for a dependence on consistent capture. The full comparison walks through each one.

Whichever you choose, the rule is the same: one method, repeated the same way, read as a trend. That’s the principle Corpus is built on. It gives you a body-fat estimate from a guided photo check-in, captured the same way every week.