Despite the fact that the arms are used extensively in daily life and that some of the muscles of the shoulder girdle share both a respiratory and a positional function for the arms, surprisingly little is known about the respiratory response to unsupported upper extremity activity. To determine the respiratory consequences of simple arm elevation during tidal breathing, we measured minute ventilation (VE), tidal volume (VT), respiratory rate (f), heart rate (HR), oxygen uptake (VO2), and carbon dioxide production (VCO2) in 22 normal subjects seated with arms elevated in front of them to shoulder level (AE) for 2 min and down at the sides (AD) for the same time period. The sequence was randomized. Compared with AD, during AE there were significant increases in VO2 (336 +/- 18 vs 289 +/- 14 ml/min, p less than 0.001), VCO2 (315 +/- 23 vs 245 +/- 16 ml/min, p less than 0.001), HR (84 +/- 6 vs 73 +/- 4 beats/min, p less than 0.05), VE (11.5 +/- 0.9 vs 9.3 +/- 0.6 L/min, p less than 0.001), and VT (868 +/- 66 vs 721 +/- 48 ml, p less than 0.001). In 11 subjects, breath-by-breath metabolic and ventilatory parameters were studied with AD for 2 min, AE for 2 min, and with AD for 3 min while also recording gastric (Pg), pleural (Ppl), and transdiaphragmatic pressures (Pdi). With AE, there was a significant increase in Pg at end inspiration (PgI, 15.4 +/- 3.2 vs 11.9 +/- 2.7 cm H2O, p less than 0.01) and in Pdi (26.5 +/- 3.4 vs 21.4 +/- 2.4 cm H2O, p less than 0.01) with no change in Pg at end expiration (PgE) or in Ppl. The increases in VO2, VCO2, VE, and VT during arm elevation persisted for 2 min after arm lowering, whereas Pgi and Pdi abruptly dropped as the arms were lowered. We conclude that simple arm elevation during tidal breathing results in significant increases in metabolic and ventilatory requirements. These increased demands are associated with higher PgI and Pdi suggesting an increased diaphragmatic contribution to the generation of ventilatory pressures. The sudden drop in Pg with arm lowering indicate a change in ventilatory muscle and or torso recruitment independent of the metabolic drive and ventilatory needs. These findings may help explain the limitation that has been reported in some normal subjects and in many patients with lung disease during unsupported upper extremity activity.