Resting metabolic rate (RMR) usually is measured by indirect calorimetry for a brief interval, and the results are then extrapolated to 24-hr resting energy production. The aim of the current study was to examine the validity of this approach by measuring the within- and between-day variability in RMR. The RMR was measured hourly in 14 healthy adults from 8:00 am to 4:00 pm after an overnight fast on two separate days. The results indicated that, during day 1, RMR remained unchanged from early morning to late afternoon, and there was no significant difference between RMR measurements 1 through 8. Based upon analysis of variance and intraclass correlation coefficients, the averages of the first and the last three RMR measurements were found to be unreliable. RMR reliability was improved by averaging all measurements taken during the day, and maximum reliability was obtained by averaging the middle three RMR measurements. A similar pattern of results was observed during day 2, although the overall trend was for the reliability of RMR measurements to improve relative to day 1. The average RMRs on days 1 and 2 were not significantly different. These results indicate that the current practice of establishing a patient's RMR based upon a single measurement potentially can lead to large errors in determining energy needs. Measurement reliability can be improved by serially measuring RMR, eliminating the initial measurement, and averaging the remaining two to three values.