PURPOSE: Glyceryl trinitrate (GTN) provoked cardioinhibitory syncope during the head-up tilt test is preceded by a period of disrupted blood pressure variability (BPV). Endogenous nitric oxide (NO) attenuates BPV independently of blood pressure (BP). We hypothesized that exogenous NO donor GTN might decrease BPV during the presyncope period. A decrease in BPV may predict the tilt outcome. METHODS: We analyzed 29 tilt test recordings of subjects with GTN-induced cardioinhibitory syncope and 30 recordings of negative subjects. A recursive autoregressive model of BPV after GTN was performed; powers of the respiratory (0.15-0.45 Hz) and nonrespiratory frequency (0.01-0.15 Hz) bands were calculated for each of the 20 normalized time periods. The post-GTN relative changes in heart rate, BP, and BPV were calculated. RESULTS: In the syncope group, spectral power of nonrespiratory frequency systolic and diastolic BPV progressively felt for 30% after GTN application and stabilized after 180 s. BP started to fall 240 s after the GTN application. Decrease in nonrespiratory frequency power of diastolic BPV 20 s after GTN administration predicted cardioinhibitory syncope (area under the curve 0.811; 77% sensitivity; 70% specificity; cutoff value > 7%). CONCLUSION: GTN application during the tilt test attenuates systolic and diastolic nonrespiratory frequency BPV during the presyncope period, independent of BP. A decrease in nonrespiratory frequency diastolic BPV 20 s after GTN application predicts cardioinhibitory syncope with good sensitivity and moderate specificity.