- Aims: Previous studies indicated that fluctuations in photoplethysmograph (PPG) amplitude may predict coronary artery disease (CAD). The respiratory modulation response (RMR) is derived from spectral analysis of the PPG signal during instructed breathing. We sought to evaluate the RMR obtained from the fingertip as a predictor of significant CAD. Methods and results: RMR was calculated as the relative change in the respiratory modulation of the PPG, in response to breathing at 0.1Hz. RMR results of 97 consecutive patients (age 62±12 years, 77% male) were compared with their angiograms. Coronary lesions with diameter stenosis >50% that required revascularisation were classified significant. RMR was analysed after 20 sec. of spontaneous breathing, followed by 70 sec. of breathing at 0.1Hz. The test was repeated post procedure in 76 patients. RMR was lower in patients with significant CAD compared to those with non-significant (16.3±20.1 ; n=66 vs. 40.6±16.9; n=31, P<0.001). It improved after successful angioplasty (from 15.0±19.0 to 40.0±18.9; P<0.001) and did not change after diagnostic catheterisation. Using receiver operating characteristic analysis, we identified RMR<30% (sensitivity 79%, specificity 87%; positive predictive value 93%; negative predictive value 66%) to be the optimal cutoff for predicting significant CAD. Conclusions: The RMR is a novel, non-invasive parameter for the assessment of significant CAD.