Mifflin-St Jeor vs. Harris-Benedict: Do the BMR Formulas Actually Differ Much
Mifflin-St Jeor is generally considered more accurate for most modern populations and tends to produce a slightly lower BMR estimate than the older Harris-Benedict formula, which was developed nearly a century ago on a narrower population. For most people, the two formulas differ by roughly 3-10%, small compared to the uncertainty introduced by activity-level self-reporting.
Both formulas estimate the same underlying number — resting metabolic rate — from similar inputs, but they were developed decades apart on different populations, which explains most of the gap between them.
Where each formula comes from
Harris-Benedict was developed in 1919 (later revised in 1984) based on a relatively small study population. Mifflin-St Jeor, published in 1990, was developed using a larger, more contemporary sample and is generally regarded by dietitians and researchers as more accurate for most people today, which is why it's the more commonly used default in modern calculators.
How much the results actually differ
For most people, the two formulas land within roughly 3-10% of each other — a real but modest gap, especially compared to the much larger uncertainty introduced by activity-level self-reporting, which is typically the single biggest source of error in a full TDEE estimate.
Which one to actually use
Mifflin-St Jeor is the generally preferred default for most adults today. Katch-McArdle, a third formula that factors in body fat percentage directly rather than estimating from weight and height alone, can be more accurate for people with a known, accurately measured body fat percentage — particularly athletes or anyone with an atypical muscle-to-fat ratio.