Hyperoxia causes vasoconstriction in most tissues, by mechanisms that are not fully understood. We investigated microvascular effects of breathing 100% oxygen in healthy volunteers, using iontophoresis to deliver acetylcholine (ACh) and sodium nitroprusside (SNP). Aspirin and vitamin C were used to test for involvement of prostaglandins and radical oxygen species. Forearm skin perfusion was measured using laser Doppler perfusion imaging. Results were analysed using dose-response modelling. The response to ACh was reduced by 30% during oxygen breathing compared to air breathing [0.98 (0.81-1.15) PU vs. 1.45 (1.30-1.60) PU, p andlt; 0.001]. ED50 values were unchanged [2.25 (1.84-2.75) vs. 2.21 (1.79-2.74), not significant]. Aspirin pre-treatment abolished the difference in response between oxygen breathing and air breathing [maximum: 1.03 (0.90-1.16) vs. 0.89 (0.77-1.01), not significant; ED50: 1.83 (1.46-2.30) vs. 1.95 (1.65-2.30), not significant]. ACh-mediated vasodilatation during 100% oxygen breathing was partially restored after pre-treatment with vitamin C. Breathing 100% oxygen did not change the microvascular response to SNP [1.45 (1.28-1.62) vs. 1.40 (1.26-1.53), not significant]. These results favour the hypothesis that hyperoxic vasoconstriction is mediated by inhibition of prostaglandin synthesis. Radical oxygen species may be involved as vitamin C, independently of aspirin, partially restored ACh-mediated vasodilatation during hyperoxia.