Despite enormous rates of minute ventilation (Ve) in the galloping Thoroughbred (TB) horse, the energetic demands of exercise conspire to raise arterial Pco(2) (i.e., induce hypercapnia). If locomotory-respiratory coupling (LRC) is an obligatory facilitator of high Ve in the horse such as those found during galloping (Bramble and Carrier. Science 219: 251-256, 1983), Ve should drop precipitously when LRC ceases at the galloptrot transition, thus exacerbating the hypercapnia. TB horses (n = 5) were run to volitional fatigue on a motor-driven treadmill (1 m/s increments; 14-15 m/s) to study the dynamic control of breath-by-breath Ve, O(2) uptake, and CO(2) output at the transition from maximal exercise to active recovery (i.e., trotting at 3 m/s for 800 m). At the transition from the gallop to the trot, Ve did not drop instantaneously. Rather, Ve remained at the peak exercising levels (1,391 +/- 88 l/min) for approximately 13 s via the combination of an increased tidal volume (12.6 +/- 1.2 liters at gallop; 13.9 +/- 1.6 liters over 13 s of trotting recovery; P < 0.05) and a reduced breathing frequency [113.8 +/- 5.2 breaths/min (at gallop); 97.7 +/- 5.9 breaths/min over 13 s of trotting recovery (P < 0.05)]. Subsequently, Ve declined in a biphasic fashion with a slower mean response time (85.4 +/- 9.0 s) than that of the monoexponential decline of CO(2) output (39.9 +/- 4.7 s; P < 0.05), which rapidly reversed the postexercise arterial hypercapnia (arterial Pco(2) at gallop: 52.8 +/- 3.2 Torr; at 2 min of recovery: 25.0 +/- 1.4 Torr; P < 0.05). We conclude that LRC is not a prerequisite for achievement of Ve commensurate with maximal exercise or the pronounced hyperventilation during recovery.