How to measure progress: strength, size and the scale

If the scale isn't moving, check your lifts first. If the weight or reps on an exercise keep going up over the weeks, you're getting stronger. Add a tape measure around your waist. And judge your body weight by its average over a few weeks, not by one morning.

Why isn't the scale enough? It's the easiest number to get, but it bounces around from day to day.^[1] Your morning weight can move by about half a kilo from one day to the next with nothing real changing.^[1] Lifting can also take off a little fat and add a little muscle.^[2][3] So the scale can show much less change than your body went through.

The last article covered how fast you can build muscle. This one is about how to tell whether it's happening.

Why the scale alone misleads you

If you judge progress by one number that bounces around, you'll make bad calls. Say you weigh 70 kg on Monday and 70.6 kg on Thursday. You might decide the program isn't working, cut your food, or quit. None of that would rest on anything real.

The fix costs almost nothing. You already write down your sets, or you should. Add a waist measurement every couple of weeks and a few morning weigh-ins a week. Now you have three signals instead of one.

What the research says

Nobody has tested these signals against each other in beginners who lift. So this section pulls together separate evidence on how much each one wobbles, plus one large study on whether keeping track helps at all.

The scale moves when nothing has changed

In one study of 65 active men, first-morning body weight varied by about half a kilo a day (0.51 kg on average).^[1] Limited evidence. The men exercised in the heat each day and drank to replace their sweat, so your own swings may be bigger or smaller. The energy balance article covers where those swings come from.

Lifting can also change your body without moving the scale much. A meta-analysis pools the results of many studies on the same question into one estimate. One from 2022 pooled 54 trials in healthy adults, women and men. Compared with not exercising, lifting for at least four weeks took off about half a kilo of fat (0.55 kg).^[2] Moderate evidence. A 2020 meta-analysis of 111 studies found that lifting added about 1.5 kg of muscle mass (1.53 kg).^[3] Moderate evidence, in healthy adult men only.

Put those side by side and you get an illustration, not a finding. Lose half a kilo of fat and gain one and a half of muscle, and the scale ends up about a kilo higher over months. A couple of ordinary daily swings can hide that. The two numbers come from separate studies of different people, so don't read them as one result.

Strength numbers repeat well

A one-rep max (1RM) is the most weight you can lift once with good form. A systematic review collects every study on a question and reads them together. One from 2020 gathered 32 studies of 1,595 people who tested their 1RM more than once. The typical difference between test days was about 4% (a median of 4.2%).^[4] Retests were reliable in trained and untrained people, men and women, and younger and older adults. Untrained people varied a little more (about 5.5% versus 3.3%).^[4] Moderate evidence.

Now the arithmetic. On a 100 kg lift, 4% is 4 kg. So a change of a kilo or two in your max can be noise.

You don't have to test a true max to use this. You can estimate one from a normal set, and the estimate is closest when the set is heavy. In a study of 70 adults (34 men and 36 women, aged 18 to 69), sets of 5 reps predicted the max best. The authors advised using no more than 10 reps.^[5] A study of 58 college football players found that predictions from 3 to 5 reps beat those from more than 6.^[6] Limited evidence. We only have a few studies on these formulas, so treat this as a best guess.

No formula fits everyone. A 2024 analysis pooled 952 tests from 269 studies, covering 7,289 people. It looked at how many reps people could do at a given share of their max.^[7] At 80% of max, people's rep counts typically spread about 2.5 reps either side of the average. At 60%, the spread was about 4.4 reps.^[7] Moderate evidence. In plain words: two people can lift 80% of their max for quite different numbers of reps. With lighter weights, the gap grows.

That's why an estimated max works best against your own past numbers. In the football study, training didn't make the estimates less accurate. But to track the change well, the test weight had to be fairly heavy.^[6] Limited evidence. Comparing the same exercise at a similar number of reps goes a step further. This is coaching practice, not a tested finding.

Put it together and strength is the signal you're already logging. No study has ranked these signals against each other. Coaching practice.

A tape measure sees what the scale misses

In 2020, an international panel on waist size wrote that waist size goes down with exercise or diet changes, with or without weight loss.^[9] Expert consensus. This is the panel's view, not a tested finding. Most of the evidence behind it comes from adults with overweight.

You can measure yourself well enough. People who measured their own waist came out about 1 to 3 cm smaller than a trained measurer did.^[9] So a self-measure is fine for tracking, as long as you do it the same way every time. That last part is coaching practice.

Arms work too. In 67 young men who trained for 10 to 12 weeks, a tape around the flexed upper arm picked up growth (+4.9%).^[10] So did an ultrasound scan, which uses sound waves to measure muscle thickness (+3.9%). Both caught the change, but the sizes differed.^[10] Limited evidence: young men, upper arm only.

A tape measures everything under it: muscle, fat and skin. So the pattern to look for is a bigger arm with a smaller waist. Coaching practice.

Body-fat scales are rough

Home body-fat scales send a small electrical current through your body and estimate fat from how it travels. Water, temperature and where your skin touches the pads all shift the number.^[11] Expert consensus. This is the authors' view, not a tested finding.

One lab study tested 15 of these devices (14 consumer models) on 73 adults: 38 women and 35 men. Everyone was tested fasted, rested and after using the toilet, so this was the best case.^[8] A "point" here is one percentage point of body fat, like 20% versus 21%. The researchers compared single readings with a detailed lab reference. Depending on the device, they were off by an average of 3.5 points too low to 11.7 points too high.^[8] On average, only five of the 15 devices matched the lab within 2 points, and one of those five was the lab-grade model.^[8] Limited evidence: one well-run study.

The researchers tested 37 of the same people again about three months later. The scales tracked changes better than they measured body fat outright. Their error was about 2 to 3 points (1.9 to 2.9).^[8] But the real average change over that time was only 0.2 points.^[8] The authors worked out that a change of 1.4 points was beyond the scale's own error. But that came from two readings taken one right after the other. They say readings on different days would almost certainly be off by more.^[8] Limited evidence.

So a body-fat scale is rough. One reading means little. A long-term trend, taken under identical conditions, might be worth a glance.

Lab scans have limits too. A DXA scan is an X-ray scan that estimates fat, muscle and bone. Even DXA needs strict conditions to repeat well: rested, fasted overnight, in minimal clothing.^[12] Food, exercise and fluids before a scan all shift the result.^[11] Expert consensus.

Keeping track helps, a little to moderately

A 2016 meta-analysis pooled 138 trials. People who were prompted to monitor their progress got closer to their goals than people who weren't.^[13] The authors call the effect small to medium. After a correction for small studies that may be missing, it looked closer to small.^[13] Moderate evidence.

Two details fit lifting well. The effect was bigger when people physically recorded their progress than when they didn't. It was also bigger when they reported it to someone or made it public, rather than keeping it private.^[13] Treat that as a strong hint, not a separate finding.

And tracking a behaviour mostly changed that behaviour, while tracking an outcome mostly changed that outcome.^[13] Limited evidence: those were smaller groups of studies. Your training log tracks the behaviour (sets and reps). Your measurements track the outcome.

How to measure progress, step by step

Everything in this section is common coaching practice. It hasn't been tested head-to-head.

Track strength

1. Log every working set. A working set is one you do for real, after the warm-up. Write down the weight and the reps.
2. Pick three or four main lifts to watch. A squat, a press, a pull and a hinge cover most of the body.
3. Compare like with like. Same exercise, similar reps. A 5-rep set and a 15-rep set give different estimates.
4. Lean on sets of 10 reps or fewer. Sets of 5 or fewer track change best. An estimate from a set of 15 or 20 reps is rougher.
5. Judge over weeks, not one session. A dip of a kilo or two on one day can be noise.

A common formula, called Epley, estimates your max as the weight × (1 + reps ÷ 30). Here's how that works over two months. Say your log reads 50 kg × 8 on the squat. That puts your estimated max at about 63 kg (50 × 1.27). Eight weeks later you squat 57.5 kg × 8, an estimate of about 73 kg. That's about 10 kg up, roughly four times the typical 4% wobble on a tested max (about 2.5 kg). An estimate from a set of 8 wobbles more, so trust a rise that holds over several weeks.

Measure your waist

1. Measure every two weeks. Do it in the morning, after the toilet and before you eat.
2. Use one spot every time. Halfway between your lowest rib and the top of your hip bone works. So does the top of the hip bone. Neither is the right one. Pick one and stick to it.
3. Measure twice and write down the average.
4. Add your arms if you like. Use the same arm, flexed, at its widest point.

Weigh yourself, then average

1. Weigh yourself several mornings a week. Same conditions each time: after the toilet, before food or drink, on the same scale.
2. Compare weekly averages, not single days.
3. If you use a body-fat scale, follow the study authors' rules. Same device, first thing in the morning, after the toilet, before eating or drinking. Look at the average over several days.^[8]

Here's step 2 with numbers. A 90 kg lifter weighs in five mornings: 90.4, 89.8, 90.6, 90.1 and 90.3. That averages 90.2 kg. Four weeks later the mornings read 90.0, 90.5, 90.2, 89.9 and 90.4. Still 90.2 kg. Flat. But the waist is 1 cm smaller and the squat estimate is up 5 kg. That's progress the scale didn't show.

Common mistakes

• Reading one weigh-in as a verdict. A jump overnight feels like fat. One morning can move half a kilo with nothing real changing.^[1] Compare weekly averages.
• Treating an estimated max as your real max. The number looks exact, so people quote it. It's an estimate, and it differs between people. Compare it only with your own past numbers on the same lift.
• Trusting a body-fat reading. The scale shows a precise-looking number, so it feels true. One reading can be off by several points. Watch the tape instead.
• Moving the tape around. Measuring at the belly button one week and the hip bone the next makes up changes. Pick one spot and keep it.
• Calling a stall after one bad session. One flat day feels like the end of progress. Judge over several weeks. If a lift stops moving for weeks, see what to do about a training plateau.

How Gymbers fits

Strength is the signal you're already logging. In the Gymbers app, each exercise has its own history, charts and records: best estimated 1RM, heaviest weight and most volume. The app estimates your one-rep max from your logged sets with the Epley formula. It does this at any rep count, so give more weight to your heavier sets. You can also log body weight, body-fat percentage and tape measurements, each with a chart and history.