The history of BMR equations
The Harris-Benedict equations were published in 1919 by Francis Gano Harris and James Arthur Benedict of the Carnegie Institution. They were the first scientifically validated formulas for predicting Basal Metabolic Rate from body measurements. The equations remained the clinical standard for over 70 years.
In 1984, Roza and Shizgal revised the original equations using a larger, more diverse sample, correcting for some of the original's systematic overestimation. In 1990, Mifflin and St Jeor published a new equation that became the modern gold standard, validated by the American Dietetic Association (now Academy of Nutrition and Dietetics) as the most accurate for most adults.
The three formulas compared
| Formula | Male BMR | Year | Accuracy |
|---|---|---|---|
| Original Harris-Benedict | 66.47 + (13.75 × kg) + (5.00 × cm) − (6.76 × age) | 1919 | ±15% |
| Revised Harris-Benedict | 88.36 + (13.40 × kg) + (4.80 × cm) − (5.68 × age) | 1984 | ±12% |
| Mifflin-St Jeor | (10 × kg) + (6.25 × cm) − (5 × age) + 5 | 1990 | ±10% |
Why the original Harris-Benedict overestimates
The original 1919 Harris-Benedict equation was developed on a relatively small, non-representative sample of primarily young, lean, white American males. Modern population studies show it systematically overestimates BMR by 5–15%, particularly for overweight and obese individuals, women, and older adults. This matters practically: if your BMR estimate is 150 kcal/day too high, your "maintenance" calories are actually a 150 kcal/day surplus — causing gradual weight gain even at supposed maintenance.
When to use Harris-Benedict specifically
The Mifflin-St Jeor equation is the recommended choice for most people. However, Harris-Benedict remains useful in specific contexts: clinical nutrition research comparing historical data, academic papers that reference the original formula, and cases where you want to see the range of estimates across multiple validated equations rather than relying on a single number.
Our Harris-Benedict Calculator shows all three equations side by side, giving you a consensus estimate and flagging when the formulas diverge significantly — which often indicates unusual body composition requiring real-world calorie tracking to calibrate.
The practical reality: all formulas are estimates
Every BMR equation has ±10–15% accuracy at the individual level. Two people with identical age, weight, height, and sex can have BMRs that differ by 300+ kcal/day due to differences in muscle mass, organ size, thyroid function, and individual metabolic variation. This is why tracking actual food intake and body weight changes over 2–3 weeks provides more reliable individual maintenance data than any formula.