BACKGROUND: To analyze the association between a 34-week military training on body composition, physical fitness and compensatory changes in resting energy expenditure (REE) recognized as adaptive thermogenesis (AT). We also explored if regional body composition changes were related to AT. METHODS: Twenty-nine male army cadets, aged 17 to 22 years were tested at baseline (T0) and after 34-weeks military training (T1). Physical training was performed 5 days/week during 90 minutes/day. Measurements included body composition by dual-energy x-ray absorptiometry; physical fitness by 3000-m running, pull-up, 50-m freestyle swimming, push-up and sit-up tests; REE measured by indirect calorimetry (REEm) and predicted from fat-free mass (FFM), fat mass (FM) and ethnicity at T0 (REEp). %AT was calculated using values at T1: 100(REEm/REEp-1); and AT (kcal/day) as %AT/100 multiplied by baseline REEm. RESULTS: Physical training was associated with increases of lean soft tissue (LST) (∆1.2±1.3 kg), FM (∆1.4±1.3 kg), FFM (∆1.2±1.3 kg) and physical fitness (P<0.01), but no REE changes (∆59.6±168.9 kcal/day) and AT were observed (P>0.05). Though a large variability was found, AT was partially explained by trunk LST (r2=0.17, P=0.027). Individuals showing a higher AT response demonstrated a higher trunk LST increase (∆0.8±0.7 kg, P<0.05). CONCLUSIONS: The military training increased LST, FM, FFM and physical fitness. Though no mean changes in AT occurred, a large individual variability was observed with some participants increasing REE beyond the expected body composition changes, suggesting a spendthrift phenotype. Changes of trunk LST may play an important role in the AT response observed in these individuals.