The cold face test (CFT) is a non-invasive challenge maneuver of the autonomic nervous system which activates the peripheral sympathetic and the cardiac parasympathetic nervous system and induces peripheral vasoconstriction and bradycardia. The physiology of CFT-induced bradycardia is still controversial. The heart rate decrease might result from a direct central up-regulation of cardiovagal activity or might be a secondary effect of baroreceptor activation or of changes of respiration. The purpose of this study was to analyze the origin of CFT-induced bradycardia. To evaluate the influence of respiration on bradycardia during CFT, we studied cardiac responses in 10 healthy volunteers during CFT (0-1 degrees C cold compresses for 60 s) with three different respiratory patterns: one with spontaneous and two with paced respiration (6 and 15 cycles/minute). We continuously monitored heart rate (HR), blood pressure (BP) and respiration and determined heart rate variability by assessment of coefficient of variation (CV), standard deviation (SD) and the root mean square of successive differences (RMSSD) of HR as well as low (LF) and high (HF) frequency spectra power of HR and BP. When coherence was above 0.5, we calculated the transfer function gain between HR and respiration in the HF band, as an index of respiratory sinus arrhythmia, and between HR and BP in the LF band, as an index of baroreflex sensitivity. HR decreased and BP increased significantly during the three types of CFT. The decrease of HR and the increase of BP, of time and frequency domain parameters did not differ between the three breathing patterns. Respiration, and HF and LF power of respiration did not change during CFT. The gain of the HF-transfer function between HR and respiration and the LF-transfer function gain between HR and BP increased significantly during CFT, but the increase did not differ between the three breathing patterns. The increase of the gain of both transfer functions is most likely due to an increase of vagal traffic and together with the unchanged respiratory pattern suggests that CFT-induced bradycardia is not due to baroreflex or respiratory influences, but seems to result from central vagal activation.