It has been suggested that physical training causes a slower, deeper breathing pattern at a given level of ventilation, but there is no convincing evidence to support this. We examined breathing pattern during maximal incremental exercise on a cycle ergometer in 7 males before and after 3-4 weeks of cycle endurance training (TRAINING), and in 6 males before and after a similar period of unaltered physical activity (CONTROL); all subjects were healthy and previously sedentary. After physical training there was a significant increase in peak oxygen uptake, and significant reductions in carbon dioxide output, heart rate (fHR) and minute ventilation (VI) at submaximal workloads; peak VI was significantly increased whereas peak fHR was unchanged. At matched VI levels (moderate, moderately-high, high) the TRAINING subjects' breathing pattern was not significantly altered; there was a power of at least 80% to detect a significant (> 0.30 L) increase in tidal volume (P < 0.05) at moderately high and high ventilation levels. There was no change in the CONTROL subjects' maximal exercise performance, or breathing pattern at matched VI levels, over the same period. Short-term, activity-specific physical training does not significantly affect the breathing pattern adopted by normal humans during progressive exercise.