To investigate the hypothesis that facial cooling (FC) exerts a greater influence on the cardiovascular system at lower versus higher levels of exercise, this study examined the effect of facial cooling [mean (SE): 0 (2)-degrees-C at 0.8 m . s -1 wind velocity] during 30 min low [35% maximum oxygen consumption (VO2max)] and moderate (70% VO2max) levels of cycle ergometry in the supine position. Five male subjects were assigned in random order to four exercise conditions: (1) FC at 35% VO2max (FC35), (2) no cooling (NFC35), (3) FC at 70% VO2max (FC70), and (4) no cooling (NFC70). Heart rate (f(c)), stroke volume (V(s)), and cardiac output (Q(c)) were measured at rest and every 10 min of exercise using impedance cardiography. During FC35, the change in f(c) [mean (SE)] was significantly lower (P<0.05) than NFC35 at 10 [22 (5) vs 31 (3) beats.min-11, 20 [29 (6) vs 35 (3) beats.min-1], and 30 [29 (5) vs 38 (4) beats . min - 1] min. No differences in f(c) were observed between FC70 and NFC70. Furthermore, FC had no effect on V(s) or Q(c) at either exercise intensity. However, when comparing the FC70 and NFC70 conditions, there was a significant main effect (P<0.05) in mean arterial pressure (P(a)BAR) response with cooling despite the fact that neither V(s) or Q(c) were different from the NFC70 control. The increase (P < 0.05) in the estimated change in systemic vascular resistance (solid triangle P(a)BAR-solid triangle Q(c)-1) could partly explain the relative rise in P(a)BAR at FC70. No pressor effect of cooling was observed at 35% VO2max. The results suggest that the FC condition promotes exercise bradycardia at low levels of exercise and exerts a greater pressor response during moderate exercise.