Objective The purpose of this study was to develop a new model, the Allocation and Interaction Model (AIM), to better predict human total energy expenditure (TEE) among a group of highly active humans living in a variety of natural environments. AIM estimates were tested to determine if it produces more accurate TEE predictions than the Factorial Method. Methods AIM includes metabolic cost terms for basal metabolic rate, thermoregulation, and the thermic effect of food, as well as more accurate activity cost estimations. AIM was tested using doubly labeled water and Flex‐Heart Rate (Flex‐HR)‐measured TEEs of healthy, highly active adults (N = 59) participating in National Outdoor Leadership School semester‐long courses. Data from a month‐long pilot study (N = 6) were also included. Results AIM produced TEE estimates that were not significantly different from measured energy expenditure values. Overall, AIM came within 4.1% of measured values; the Factorial Method underestimated by over 25%. At TEEs greater than 3,000 kcal day−1, AIM underestimated TEE by 11% compared to 31.6% by the Factorial Method. Also, at TEEs greater than 3,000 kcal day−1, the Flex‐HR method overestimated TEE by 17%. Conclusions This analysis demonstrated that AIM is more accurate than the Factorial Method for predicting TEE across a range of climates and physical activity levels. This suggests that AIM should be used in place of the Factorial Method for estimating human TEE. Am. J. Hum. Biol. 28:372–380, 2016. © 2015 Wiley Periodicals, Inc.