Purpose To compare energy use and substrate partitioning arising from repeated lower- versus upper-body sprints, or endurance exercise, across a 24-h period. Methods Twelve untrained males (24 ± 4 y) completed three trials in randomized order: (1) repeated sprints (five 30-s Wingate, 4.5-min recovery) on a cycle ergometer (SITLegs); (2) 50-min continuous cycling at 65% V̇O2max (END); (3) repeated sprints on an arm-crank ergometer (SITArms). Respiratory gas exchange was assessed before and during exercise, and at eight points across 22 h of recovery. Results Metabolic rate was elevated to greater extent in the first 8 h after SITLegsthan SITArms(by 0.8 ± 1.1 kJ/min, p = 0.03), and tended to be greater than END (by 0.7 ± 1.3 kJ/min, p = 0.08). Total 24-h energy use (exercise + recovery) was equivalent between SITLegsand END (p = 0.55), and SITLegsand SITArms(p = 0.13), but 24-h fat use was higher with SITLegsthan END (by 26 ± 38 g, p = 0.04) and SITArms(by 27 ± 43 g, p = 0.05), whereas carbohydrate use was higher with SITArmsthan SITLegs(by 32 ± 51 g, p = 0.05). Plasma volume-corrected post-exercise and fasting glucose and lipid concentrations were unchanged. Conclusion Despite much lower energy use during five sprints than 50-min continuous exercise, 24-h energy use was not reliably different. However, (i) fat metabolism was greater after sprints, and (ii) carbohydrate metabolism was greater in the hours after sprints with arms than legs, while 24-h energy usage was comparable. Thus, sprints using arms or legs may be an important adjunct exercise mode for metabolic health.