In a study of high-quality research methods, the Harris-Benedict equation predicted RMR within 10% of measured RMR in 64% of obese individuals, and, of the 36% remaining, 30% were overestimations and 6% were underestimations. A concurring, high quality research methods study reported that the Harris-Benedict equation predicted within +/-10% of measured RMR in 64% of obese males; the remaining 36% were overestimation errors. In a moderate quality research design, the Harris-Benedict equation predicted RMR within 10% of measured RMR in obese individuals by 39%; in the remaining 61%, 52% were overestimations and 9% were underestimations. In another moderate quality research design study with only females, 59% were within +/-10% of predicted values using Harris-Benedict; of the remaining 41%, 21% were overestimations and 20% were underestimations. A small study reported individual RMR estimation errors in overweight and obese women of approximately 396 kcals/day using actual weight. Overestimations by 43% to underestimations by 35% occurred in obese individuals using all studies that reported individual errors ranges.

In a high quality research methods study, the Harris-Benedict equation predicted RMR within +/- 10% in only 26% of obese males; and of the remaining 74%, 4% were overestimations and 70% were underestimations. A neutral research method design reports Harris-Benedict estimated within +/- 10% of measured RMR in 13% of obese individuals with the remaining 87% as underestimations. The individual error range was a maximal overestimation by 5% to an underestimation by 94%.