MyFitnessPal review: the world's food log, and its weakest link

How accurate is the database, really?

This is the question everything else hangs on, and unusually for consumer health there is a clean answer. A 2025 study in the Journal of Human Nutrition and Dietetics had two raters independently enter the same 43 three-day food records into MyFitnessPal and Cronometer, then compared both against ESHA Food Processor running the Canadian Nutrient File.

MyFitnessPal showed poor validity for total energy, carbohydrate, protein, cholesterol, sugar and fibre — with the energy, carbohydrate and sugar discrepancies driven by women's records and protein by men's. It was also inconsistent between raters for sodium and sugar. Cronometer, on the same records, showed good reliability and validity for most nutrients.

An earlier JMIR validation against the Belgian Nubel database found MyFitnessPal underestimated protein by 7.8%, carbohydrate by 6.4% and fat by 1.7% while overestimating energy by 1.3%, concluding it was adequate for energy and macros but not for cholesterol and sodium.

Why the database behaves this way

Architecture, decided a decade ago. MyFitnessPal's advantage has always been an enormous crowdsourced catalogue — millions of user-submitted entries, which is exactly why it can find the specific burrito from the specific chain in four seconds. Those entries are not verified before they become searchable.

Search a common food and you get a wall of near-duplicates whose values disagree, sometimes by a factor of two. Choosing between them is a judgement call the interface does not equip you to make, which is precisely why two trained raters produced two answers.

The practical fix, if you stay: use only entries carrying the green verified tick. It removes most of the error and roughly doubles the time it takes to log a meal. That trade is the whole review in one sentence.

An app that cannot reproduce its own answer cannot be audited — not by you, and not by the dietitian reading your log.

Does the inaccuracy actually matter?

Less than the numbers suggest, in one specific case, and more than you'd like in the rest.

For weight management, a consistently biased estimate still produces a usable trend. If your log reads 15% low every day, the direction of travel is still informative and the app still does its job. That's the honest defence and it holds.

It collapses in three places. Micronutrients — iron, B12, vitamin D, potassium — are the reason many people open a food log at all, and coverage here is not adequate. Clinical or sports contexts, where a dietitian reads your log and entry noise becomes their problem; the 2025 study exists because sports dietitians needed to know which app to trust. And sodium and cholesterol specifically, which both studies flag by name.

The paywall, and the company it keeps

Barcode scanning moved behind Premium in 2022 and remains the most-complained-about change in the category. The free tier gives you manual logging, a limited number of photo scans and a fasting timer; Premium is $79.99 a year, Premium+ $99.99. Cronometer includes barcode scanning free and tracks 84 nutrients on its free tier.

Worth knowing for context: MyFitnessPal acquired Cal AI in March 2026, and Apple removed that app from the App Store the following month over deceptive billing design — a weekly price displayed more prominently than the real charge, an obscured auto-renewal toggle, and a second purchase flow shown to users who declined the first. It was reinstated after changes. That's not MyFitnessPal's own paywall, but it is now MyFitnessPal's company, and our news report covers it in full.

Living with it for 31 days

Fast, familiar and genuinely good at removing friction. Our tester logged 29 of 31 days without effort, which is a higher adherence rate than anything else we've tested — and adherence is not a trivial virtue in a category where most people quit in week two.

We logged three dark patterns in the window: an upgrade interstitial on cold start, a countdown on an annual offer, and the barcode scanner presented as available before revealing the paywall. Category median is five, so this is mid-pack rather than egregious.

Who should use it

• Use it if speed and adherence are your bottleneck and you eat out constantly.
• Use it carefully — verified entries only — if you're tracking anything beyond calories.
• Switch to Cronometer if you track micronutrients, work with a clinician, or manage sodium or cholesterol.
• Skip if you're on a GLP-1 — the app is built to help you eat less, which the medication has handled.

The bottom line

MyFitnessPal earns 64/100. It is the best-adhered-to food log in the world and the least accurate one we've tested against a reference standard. If you only need a calorie trend, that's a fair trade. If you need a number to be true, it isn't.