Exercising muscle releases interleukin-6 (IL-6), but the mechanisms controlling this process are poorly understood. This study was performed to test the hypothesis that the IL-6 release differs in arm and leg muscle during whole-body exercise, owing to differences in muscle metabolism. Sixteen subjects (10 men and six women, with body mass index 24 ± 1 kg m(-2) and peak oxygen uptake 3.4 ± 0.6 l min(-1)) performed a 90 min combined arm and leg cycle exercise at 60% of maximal oxygen uptake. The subjects arrived at the laboratory having fasted overnight, and catheters were placed in the femoral artery and vein and in the subclavian vein. During exercise, arterial and venous limb blood was sampled and arm and leg blood flow were measured by thermodilution. Lean limb mass was measured by dual-energy X-ray absorbtiometry scanning. Before and after exercise, biopsies were obtained from vastus lateralis and deltoideus. During exercise, IL-6 release was similar between men and women and higher (P <0.05) from arms than legs (1.01 ± 0.42 and 0.33 ± 0.12 ng min(-1) (kg lean limb mass)(-1), respectively). Blood flow (425 ± 36 and 554 ± 35 ml min(-1) (kg lean limb mass)(-1)) and fatty acid uptake (26 ± 7 and 47 ± 7 µmol min(-1) (kg lean limb mass)(-1)) were lower, glucose uptake similar (51 ± 12 and 41 ± 8 mmol min(-1) (kg lean limb mass)(-1)) and lactate release higher (82 ± 32 and -2 ± 12 µmol min(-1) (kg lean limb mass)(-1)) in arms than legs, respectively, during exercise (P <0.05). No correlations were present between IL-6 release and exogenous substrate uptakes. Muscle glycogen was similar in arms and legs before exercise (388 ± 22 and 428 ± 25 mmol (kg dry weight)(-1)), but after exercise it was only significantly lower in the leg (219 ± 29 mmol (kg dry weight)(-1)). The novel finding of a markedly higher IL-6 release from the exercising arm compared with the leg during whole-body exercise was not directly correlated to release or uptake of exogenous substrate, nor to muscle glycogen utilization.