The purpose of this study was to monitor the heart rates (HR) of seven, well-trained (maximal oxygen uptake [V̇O2max] 5.0 ± 0.5 l·min−1), competitive cyclists during a 4-d cycle stage race. On consecutive days, subjects competed in a 16.0-km individual time trial (TT), a 110.0-km mass-start road race (RR1), a 5.5-km individual hill climb (HC), and a 105.0-km mass-start road race (RR2). Within 10 d of the final race, cyclists underwent a test to determine V̇O2max, peak power output, and maximal HR. Comparison of the HR responses to each race revealed that the individual events were performed at a relatively high and constant work rate (91.1 ± 2.5% and 93.2 ± 4.7% of the maximal HR as measured in the field (HRmax) for the TT and HC, respectively). In contrast, despite similar racing speeds (42.2 ± 1.0, 39.9 ± 0.2, and 40.6 ± 0.5 km·h−1 for the TT, RR1, and RR2, respectively), the HR responses to the longer mass-start races were reduced to 81.9 ± 9.6% and 78.6 ± 8.9% of HRmax and were random in frequency and amplitude. Such stochastic changes in HR were seemingly unrelated to course terrain but may be due to the group dynamics of the cyclists. The results of this study reveal the stochastic nature of bunch cycle racing and show that the HR responses of competitive cyclists are more a function of tactical bunch riding than of terrain.