Oxidative stress promotes vascular dysfunction in chronic kidney disease (CKD). We utilized the cutaneous circulation to test the hypothesis that reactive oxygen species derived from NADPH oxidase and xanthine oxidase impair nitric oxide (NO)-dependent cutaneous vasodilation in CKD. Twenty subjects, 10 stage 3 and 4 patients with CKD (61 ± 4 yr; 5 men/5 women; eGFR: 39 ± 4 ml·min(-1)·1.73 m(-2)) and 10 healthy controls (55 ± 2 yr; 4 men/6 women; eGFR: >60 ml·min(-1)·1.73 m(-2)) were instrumented with 4 intradermal microdialysis fibers for the delivery of 1) Ringer solution (Control), 2) 10 μM tempol (scavenge superoxide), 3) 100 μM apocynin (NAD(P)H oxidase inhibition), and 4) 10 μM allopurinol (xanthine oxidase inhibition). Skin blood flow was measured via laser-Doppler flowmetry during standardized local heating (42°C). N(g)-nitro-l-arginine methyl ester (L-NAME; 10 mM) was infused to quantify the NO-dependent portion of the response. Cutaneous vascular conductance (CVC) was calculated as a percentage of the maximum CVC achieved during sodium nitroprusside infusion at 43°C. Cutaneous vasodilation was attenuated in patients with CKD (77 ± 3 vs. 88 ± 3%, P = 0.01), but augmented with tempol and apocynin (tempol: 88 ± 2 (P = 0.03), apocynin: 91 ± 2% (P = 0.001). The NO-dependent portion of the response was reduced in patients with CKD (41 ± 4 vs. 58 ± 2%, P = 0.04), but improved with tempol and apocynin (tempol: 58 ± 3 (P = 0.03), apocynin: 58 ± 4% (P = 0.03). Inhibition of xanthine oxidase did not alter cutaneous vasodilation in either group (P > 0.05). These data suggest that NAD(P)H oxidase is a source of reactive oxygen species and contributes to microvascular dysfunction in patients with CKD.