Wide variations in respiratory rate and hypoxic stimulation of chemoreceptors may produce unreliable autonomic results in patients with COPD. We studied the reproducibility of two consecutive measurements of heart rate variability (HRV) and blood pressure variability (BPV) by time frequency analysis in patients with COPD while controlling respiratory rate and oxygen hemoglobin saturation (SaO(2)). Reproducibility was assessed by paired t-tests and correlation analyses between repeated measures. Correlation analyses of the log transformed low (LF) and high frequency (HF) HRV were x 11.5 +/- 1.1 in measurement A and x 11.5 +/- 1.0 in measurement B (r = 0.89, p < 0.0001), and x 10.5 +/- 1.1 in measurement A and x 10.6 +/- 1.1 in measurement B (r = 0.89, p < 0.0001) respectively. The log transformed LF and HF BPV were x 4.9 +/- 1.3 in measurement A and x 5.3 +/- 0.9 in measurement B (r = 0.70, p < 0.0002), and x 6.4 +/- 1.3 in measurement A and 6.6 +/- 1.2 in measurement B (R = 0.71 p < 0.0001) respectively. In conclusion, time frequency analysis of HRV and BPV is reproducible and reliable in patients with COPD while controlling their respiratory rate and oxygen hemoglobin saturation. Reproducibility of these measurements may allow for a non-invasive evaluation of autonomic tone in response to treatments in COPD patients.