A smart scale reads weight to 50 grams and body fat to within 3–5%
In short
Home smart scales measure body weight very accurately — a premium model with four load sensors reads to 50 grams, roughly 0.05% variance for a 90kg user. Body fat percentage is a different matter: it is estimated from the resistance of a small current passed foot to foot, and even those premium models land around 3–5 percentage points off. The 13 or 45 metrics on screen are not separate measurements but values derived from weight and resistance, and claims like 98% agreement with DEXA usually come from the manufacturer's own testing. For a lifter, the most reproducible progress metric is not a number on the scale but the Big 3 log.
A smart scale displays dozens of metrics, but it measures two: the weight pressing on the platform, and the resistance a small current meets travelling foot to foot through the body. Everything else is computed from those two. Once that split is clear, which numbers to trust sorts itself out.
Why is the weight reading so good?
Weight is physically measured, so it is accurate. A premium model with four load sensors reads to 50 grams, about 0.05% variance for a 90kg user. There is no reason to doubt the weight itself. The problem is that the number says nothing at all about what the weight is made of.
How far can body fat percentage be trusted?
Body fat is estimated, not measured. Fat and lean tissue conduct differently, and that difference is fed into a regression equation. Even on units rated best in class, body fat accuracy sits around 3–5 percentage points. A screen reading of 18% means the real value is somewhere around 15–23%. And if it moved 1% overnight, what moved was water, not fat.
Metric counts mislead too. One scale advertises 13 metrics, another advertises 45 from eight sensors — that is not 45 measurements, it is 45 display values derived from the same raw signals. Figures like 98% accuracy versus DEXA generally come from studies the brand ran itself, so they should not be weighted like independent validation.
How should you actually weigh yourself?
- Same time, same conditions: on waking, after using the bathroom, before eating or drinking, barefoot.
- Never straight after training, after a sauna, or the morning after drinking. Hydration state changes resistance wholesale.
- Do not judge from a single day. Weigh daily but compare weekly averages.
- Read the direction on one device, not the absolute value. Switching devices ends the trend line there.
- Ignore the decimal on body fat. Inside a 3–5 point error band, a decimal is not information.
A bioimpedance scale is not a medical device for diagnosing health status. Models that add features such as an ECG reading do not replace a clinical assessment either — an abnormal reading or a symptom belongs with a doctor, not with the scale's app. Fat gain during a weight-gain phase is covered in the fat-to-muscle ratio of bulking, and sustainable eating in the diet you can actually sustain.
Why body weight is a particularly bad metric for lifters
Lifters are the group whose body weight can sit still while the body changes completely. Three months where the scale moved less than a kilo but the squat went up 20kg is a successful block; three months where body weight dropped 5kg and the bench fell 10kg is a failed one. The scale cannot tell those two apart. It reports the first as no change and the second as success.
So what goes in the log?
Reproducibility decides it. Body fat percentage carries a 3–5 point error and swings with hydration, while a 100kg squat is 100kg in any gym on any day. The axis for judging progress should be the Big 3, with body weight sitting beside it as supporting information. A bodyweight-adjusted score ties the two into one number, and brings its own trap — it rises from weight loss alone, because it corrects for body weight. In a cut, read the score and the absolute loads together.
Frequently asked questions
Is smart scale body fat percentage accurate?
Even premium models land around 3–5 percentage points off. A reading of 18% means the true value is somewhere near 15–23%, so the sensible use is the direction of change on one device rather than the absolute number.
Are the dozens of metrics each measured separately?
No. A scale measures weight and electrical resistance; muscle mass, body water, visceral fat and the rest are values derived from those two. A scale advertising 45 metrics is not taking 45 measurements.
When should you weigh yourself?
On waking, after using the bathroom, before eating or drinking, barefoot — that keeps conditions most consistent. Straight after training, after a sauna, or the morning after alcohol, hydration is shifted enough to distort the reading.
Daily or once a week?
Weigh daily and compare weekly averages. Day-to-day body weight can swing more than a kilogram on food and water alone, so judging direction from a single reading misreads it.
What should a lifter track instead of body weight?
The Big 3. Squat, bench press and deadlift 1RMs do not drift with hydration or estimation error and read the same anywhere, which makes them far more reproducible than a body fat percentage for judging progress.
Source: Men's Health