This study compared breathing frequency (fB) during a ventilation threshold (VT) test with fB during a simulated 40 km time trial (40TT) to assess whether fB can be used to monitor exercise intensity. Twenty-six male cyclists (mean +/- SD VO2max = 4.42 +/- 0274 l.min-1) performed an incremental VT ergometer test and a simulated 40TT on a set of wind-loaded rollers. During the 40TT, expired gas was collected for 3 min using open circuit spirometry at 10 km intervals. A significant correlation (r = 0.89; p < or = 0.05) was found between the fB threshold (VTf) and the gas exchange VT (VTg) (3.42 +/- 0.08 and 3.31 +/- 0.09 l.min-1, respectively). No differences were found in VO2 at the VTf vs. X VO2 (3.48 +/- 0.08 l.min-1) during the simulated 40TT. However, when fB in br.min-1 calculated from the VT test (32.5 br.min-1), was compared with fB during the 40TT (45.6 br.min-1), there was a significant difference. Ventilation, heart rate and the respiratory exchange ratio were all significantly higher during the 40TT. Although a significant relationship was found between the VT test and the 40TT for VTf measured in l.min-1 of VO2 (r = 0.66; p < or = 0.05), fB in br.min-1 obtained from the incremental test should not be used to identify exercise training intensity. These data suggest that during competition, trained athletes exercise at a higher fB than that determined when being tested, implying that fB in br.min-1 may not be a valid measurement of exercise intensity.