Daily, if the goal is catching a real trend instead of reacting to a single number. A cohort study that followed 9,768 smart scale users for an average of nearly three years found that only people who weighed in every single day showed a consistent downward trend in their weight over time. People who weighed in every other day, a few times a week, or once a week showed close to no measurable trend in either direction, not because their weight wasn't changing, but because day-to-day fluctuation buried the signal at that frequency.
The short answer
Daily weighing is the only frequency in the data that reliably shows a trend. Weighing less often doesn't mean your weight isn't changing, it means the fluctuation between one weigh-in and the next is usually bigger than the actual trend, so it gets lost in the noise. If you can tolerate a daily number without it affecting your day, daily is the more informative choice. If you can't, less frequent weighing paired with other progress signals is a reasonable tradeoff, not a lesser one.
What a 9,768-user study found
A 2021 study in the Journal of Medical Internet Research analyzed longitudinal data from Vuorinen and colleagues, tracking 9,768 Withings smart scale users across normal weight, overweight, and obese BMI groups over an average follow-up of nearly three years. The researchers grouped users by how often they weighed in and measured the average daily weight change within each group. Only the daily-weighing group showed a consistent negative trend, averaging about 0.058 kilograms of loss per day of tracking. Every other frequency group, weighing every other day, two to four times a week, or once or twice a week, showed a slight positive average change instead, essentially flat or trending up rather than down.
Same population, same scales, four different weigh-in frequencies. Only the daily group showed a real downward trend.
The correlation between weighing frequency and weight change held across all three BMI groups, though the researchers described it as modest in size even where it was statistically significant. That's an important caveat: this is observational data from people who chose their own weighing habits, not a randomized trial assigning frequency at random, so it can't fully rule out the possibility that people already losing weight successfully were also more motivated to weigh in daily. Even accounting for that, the size of the dataset and the consistency of the pattern across BMI groups makes it one of the stronger pieces of real-world evidence available on this specific question.
Why weighing less often shows no clear trend
Body weight moves by a pound or more across a single day from water, food still in digestion, and sodium intake, a swing covered in more depth in how much your weight actually fluctuates in a day. That noise is large relative to how much real fat loss happens between two points a week or two apart. Daily weighing averages that noise out over many data points, which is exactly what makes a trend visible early. Weighing once a week compares two single readings, each one still carrying that same day-to-day swing, so a real change can look identical to normal noise for weeks longer than it would with daily data.
What happens when you stop weighing altogether
The same study found that 72.5 percent of participants took at least one break from weighing longer than 30 days at some point during the study period. During those breaks, participants gained an average of 0.85 kilograms, roughly 1.9 pounds, more than they had weighed before the break started. That's not proof that skipping the scale causes weight gain on its own, but it's consistent with a simpler explanation: the scale is one of the few signals that catches a small upward trend early enough to act on it, and without that signal, small drifts have more room to compound before anyone notices.
The tradeoff daily weighing doesn't measure
None of this data captures how a person feels about the number itself. A daily weigh-in that spikes anxiety, triggers restrictive eating, or turns into checking the scale multiple times a day isn't the outcome the study is describing, and it can undo more progress than the earlier trend detection is worth. If daily weighing consistently makes your day worse before it's even started, that's a real cost worth weighing against the earlier-detection benefit. A weekly weigh-in paired with other signals, covered in what are non-scale victories, is a legitimate alternative for that situation, not a consolation prize.
How to weigh in so the number means something
Consistency matters more than precision here. Weigh in at roughly the same time each day, ideally first thing in the morning before eating or drinking, since that removes most of the same-day fluctuation from the comparison. Wearing roughly the same amount of clothing each time, or none at all, removes another small source of noise that otherwise has to be mentally subtracted from every reading. Look at the direction of the trend across a week or two rather than any single morning's reading, since a single higher number after a salty meal or a poor night's sleep tells you almost nothing on its own. A plateau that stalls for a genuinely long stretch is a different situation than ordinary noise, and worth reading about separately in why did your weight loss stall on a GLP-1.
The type of scale matters less than the habit itself. A basic digital scale that only shows total weight is enough to track the trend this research is describing, and body composition scales that estimate fat percentage or water weight add convenience but not necessarily more useful signal, since those estimates carry their own measurement noise on top of the same day-to-day fluctuation already covered above.
How OffRamp helps
OffRamp's Regain Radar reads your weigh-ins as a trend, not a verdict on any single day. That matches what the smart scale data actually shows: a daily number is only useful in aggregate, and a single reading in isolation is close to meaningless. You can log as often as you're comfortable with, and the app is built to surface the direction of the line, not to react to whichever number showed up this morning.
