Nitric oxide (NO) bioavailability is important in vascular health, but unsuitable as a clinical measure due to biological oxidation. Total nitrogen oxides (NO(x)) are stable but background nitrate levels make it difficult to detect disease-based variation. We investigated the clinical discriminatory value of NO(x) as it relates to exercise capability (VO(2peak)) and brachial artery reactivity (BAR, an NO-dependent measure of endothelial health), in healthy (H), increased risk (RF), and known cardiovascular disease (CVD) subjects. BAR was measured using forearm occlusion/hyperemia stimulus. Subjects performed a maximal graded exercise test (GXT). Blood at rest, exercise termination, and 10 min into recovery was mixed equally with 0.1 M NaOH at 4 degrees C, filtered, and stored at -70 degrees C. NO(x) was measured by chemiluminescence. Seven of the RF group then exercise-trained for 6 months prior to retesting. The H group (n = 12) was younger, had higher VO(2peak), HDL levels, and baseline NO(x) values than the RF (n = 15) and CVD (n = 10) groups. NO(x) increased from baseline to recovery in the H group only (75.85 +/- 19.04 microM vs 97.76 +/- 31.93 microM; P <or= 0.01). BAR was greater in the H versus CVD group (7.24 +/- 3.78% vs 2.59 +/- 3.53%; P <or= 0.01). The relation between VO(2peak) and NO(x) recovery was significant across groups (r = 0.71, P <or= 0.01). Following training the RF subjects (n = 7) increased VO(2peak) (29.98 +/- 6.74 to 33.08 +/- 5.57 ml kg(-1) min(-1); P <or= 0.05), BAR (3.19 +/- 3.96 to 6.86 +/- 4.81%; P <or= 0.05), and recovery NO(x) (18.29 +/- 6.46 to 66.48 + 21.11 microM; P <or= 0.01). These findings suggest that plasma NO(x) following GXT discriminate cardiovascular disease status, are related to regional endothelial function, and respond favorably to exercise training.