Are Smart Scales Accurate for Body Fat? What the Studies Say

For body weight, usually yes. For body fat, not really, at least not from a single reading. A study that compared three consumer smart scales with DXA, a scan often used as the reference method in research, found the scales measured weight well but underestimated fat mass, and the gap for an individual could be several kilograms ^[1]. The authors concluded the body-composition estimates were not accurate enough for individual clinical use ^[1].

That does not make the feature useless. Used carefully, a smart scale's body-fat figure is an estimate that can show a direction over weeks. It is not a measurement to base decisions on from one morning to the next.

This guide covers how these scales work, what studies found, why readings bounce around, and a simple weekly routine that makes the trend more trustworthy.

Most smart scales use bioelectrical impedance analysis, or BIA. When you stand barefoot on the metal pads, the scale sends a very small signal through your body and measures how easily it passes. Water-rich tissue and fat respond differently, so the scale feeds that signal into an equation to estimate how much of you is fat and how much is lean.

Two things follow from that:

• The scale never sees fat directly. It infers it using the manufacturer's own equations, which were built on groups of people who may not resemble you.
• Water changes the reading. Anything that shifts how much water is in your body can move the body-fat estimate, even when your actual fat has not changed.

It also helps to know what the reference method reports. DXA gives a figure for "lean mass," which includes water, organs and connective tissue as well as muscle. So even the lab scan is not a pure muscle count, and a bathroom scale's "muscle" number is a further step removed.

Always follow the safety warnings in your scale's manual, including any about implanted medical devices or pregnancy.

Several studies have compared BIA devices with DXA in different groups of people. The details differ, but the direction is consistent.

A pattern runs through these results. BIA can be reasonable for the average of a group, which is how researchers often use it, but the error for any one person can be large ^[4]. When you step on your own scale, that individual error is the one that matters.

The finding that error grows at higher body fat ^[2] is worth knowing if you are early in a weight-loss journey, including on GLP-1 medication. The number on the screen may be further from the truth for you than for a leaner person, and the size of that gap may change as your body changes.

None of this means the scale is lying about your weight. It means the body-fat percentage is a best guess wrapped in a precise-looking number.

Because BIA relies on the water in your body, everyday changes in hydration can show up as changes in "body fat." The main things that move the reading are:

• Time of day. A reading taken in the evening may not match one taken first thing in the morning.
• Food and drink. A large meal or a lot of fluid shortly before weighing can change the estimate.
• Exercise. A recent workout changes hydration and can nudge the number either way.
• The scale itself. Each brand uses its own equations, so two scales can give different body-fat figures for the same person on the same morning.

So a body-fat reading that drops overnight is unlikely to be real fat loss, and a jump after a big dinner is unlikely to be real fat gain.

Real change is slow by comparison. The CDC notes that people who lose weight at a gradual, steady pace of about 0.5–1 kg (1–2 lb) a week are more likely to keep it off ^[7]. Only part of that is fat, so a genuine week-to-week change in fat can easily be smaller than the day-to-day wobble in the estimate. Averaged over several weeks, under the same conditions, the real direction has a much better chance of showing through.

You cannot make a smart scale more accurate, but you can make it more consistent. Consistency is what turns noisy readings into a usable trend.

The weekly routine

1. Use one scale. Readings from different brands are not comparable. If you replace your scale, treat the new one as a fresh starting point.
2. Put it in one spot. Keep it on the same hard, flat floor, not on carpet, and do not move it around the house.
3. Weigh at the same time, in the same state. First thing in the morning works well for most people: after using the bathroom, before eating, drinking or exercising.
4. Stand the same way. Bare feet, same clothing or none, feet placed the same way on the pads.
5. Keep your profile up to date. Check that the height, age and sex in the scale's app are correct, since its equations use them.
6. Weigh on a set schedule. Either weigh every morning and use the weekly average, or weigh on the same few mornings each week.
7. Compare weeks, not days. Look at this week's average against last week's, and judge the direction over several weeks.
8. Note anything unusual. Illness, travel, a very late meal or a hard workout the evening before can all explain an odd reading.

How to read what you see

• Weight trending down, body-fat estimate trending down or flat, strength holding or improving: generally an encouraging picture.
• Weight trending down but strength steadily falling: worth a closer look, whatever the body-fat number says.
• Body-fat estimate moving the wrong way for a single week: usually noise. Wait for the trend.

If a reading worries you, a single number from a bathroom scale is not a diagnosis. Talk to your doctor or clinician, who can look at the full picture.

Many people buy a smart scale because they want to know what their weight loss is made of. That is a fair question. In the STEP 1 body-composition sub-study, of an average 13.6 kg (30 lb) lost on semaglutide, about 5.3 kg (11.7 lb) was lean mass ^[5]. In the SURMOUNT-1 sub-study with tirzepatide, roughly 75% of the weight lost was fat mass and 25% was lean mass ^[6]. A noisy body-fat reading on its own is not enough to answer that question for you.

A better approach is to read three signals together:

• Your weight trend. The weekly average, not the daily number.
• Your strength. Whether the weights or repetitions in your regular strength sessions are holding steady or going up. Muscle-strengthening activity on at least 2 days a week is the WHO recommendation for adults ^[8].
• Your body-fat estimate trend (optional). From the same scale, in the same conditions, over several weeks.

When these agree, you can be more confident in the picture. When they disagree, trust the strength and weight trends more than the body-fat estimate, and give it a few more weeks. For more on what to look for, see signs you are losing muscle, not fat and how to lose fat without losing muscle. If protein is part of your plan, the protein calculator gives a general, evidence-based range.

Dyadfit is being built around this kind of weekly check, reading your weight trend, strength progress and, if you have one, your smart scale's trend together rather than relying on any single number.