Acute systemic hypoxia causes significant increases in human skeletal muscle sympathetic nerve activity (MSNA), heart rate and ventilation. This phenomenon is thought to be primarily mediated by excitation of peripheral chemoreceptors sensing a fall in arterial free oxygen partial pressure (Pa,O2). We directly tested the role of Pa,O2 on MSNA (peroneal microneurography), heart rate, ventilation and leg haemodynamics (n = 7-8) at rest and during rhythmic handgrip exercise by using carbon monoxide (CO) to mimic the effect of systemic hypoxia on arterial oxyhaemoglobin (approximately 20 % lower O2Hba), while normalising or increasing Pa,O2 (range 40-620 mmHg). The four experimental conditions were: (1) normoxia (Pa,O2 approximately 110 mmHg; carboxyhaemoglobin (COHb) approximately 2 %); (2) hypoxia (Pa,O2 approximately 40 mmHg; COHb approximately 2 %); (3) CO + normoxia (Pa,O2 approximately 110 mmHg; COHb approximately 23 %); and (4) CO + hyperoxia (Pa,O2 approximately 620 mmHg; COHb ~24 %). Acute hypoxia augmented sympathetic burst frequency, integrated MSNA, heart rate and ventilation compared to normoxia over the entire protocol (7-13 bursts min-1, 100-118 %, 13-17 beats min-1, 2-4 l min-1, respectively, P < 0.05). The major new findings were: (1) CO + normoxia and CO + hyperoxia also elevated MSNA compared to normoxia (63-144 % increase in integrated MSNA; P < 0.05) but they did not increase heart rate (62-67 beats min-1) or ventilation (6.5-6.8 l min-1), and (2) despite the 4-fold elevation in MSNA with hypoxaemia and exercise, resting leg blood flow, vascular conductance and O2 uptake remained unchanged. In conclusion, the present results suggest that increases in MSNA with CO are not mediated by activation of the chemoreflex, whereas hypoxia-induced tachycardia and hyperventilation are mediated by activation of the chemoreflex in response to the decline in Pa,O2. Our findings also suggest that Pa,O2 is not an obligatory signal involved in the enhanced MSNA with reduced blood oxygenation.