Experimental and clinical studies have suggested atherosclerotic effects of uric acid (UA) on vascular smooth muscle cells (VSMCs). Nitroglycerine-induced vasodilation (NID), a control test for flow-mediated vasodilation, can be used as a possible marker of VSMC dysfunction. However, there is little information on the association between UA and NID. Therefore, we investigated the association between serum UA levels and NID according to sex. We measured NID of the brachial artery in 598 women (mean age: 66.2 ± 12.0 years) and 1008 men (mean age: 59.0 ± 18.0 years). In women, the mean serum UA level was 5.06 ± 1.24 mg/dL. Serum UA levels were negatively correlated with NID (p < 0.001), and NID significantly decreased with increasing serum UA levels (≤4.0 mg/dL, 13.4 ± 6.4%; 4.0 to ≤5.0 mg/dL, 11.4 ± 5.3%; 5.0 to ≤6.0 mg/dL, 10.8 ± 5.7%; >6.0 mg/dL, 9.7 ± 5.7%; p < 0.001). The prevalence of VSMC dysfunction, defined as NID < 8.4%, the division points for the lowest and middle tertiles of NID in women, increased with increasing serum UA levels ( ≤ 4.0 mg/dL, 23.3%; 4.0 to ≤5.0 mg/dL, 30.9%; 5.0 to ≤6.0 mg/dL, 36.4%; >6.0 mg/dL, 44.6%; p < 0.001). Multiple logistic regression analysis showed a significant association between serum UA levels and VSMC dysfunction (odds ratio, 1.21; 95% confidence interval, 1.02─1.43; p = 0.03). There was no interaction between age (<50 or ≥50 years) and the effect of serum UA levels on VSMC dysfunction (p interaction = 0.88). In contrast, no association was observed between serum UA levels and NID in men. Serum UA levels were significantly associated with VSMC dysfunction as assessed by NID in women.